Method and system for optimizing logistics hub locations for multi-line collaborative production

Abstract

This invention proposes a method and system for optimizing warehouse locations in logistics hubs for multi-production line collaborative production, belonging to the field of manufacturing logistics and warehousing optimization. The method includes: constructing a logistics hub warehouse location optimization model for multi-production line collaborative production; performing multi-objective collaborative optimization using the NSGA-II algorithm based on the logistics hub warehouse location optimization model; outputting a Pareto optimal solution set representing the trade-offs between different objectives; performing simulation verification and sensitivity analysis on the obtained Pareto optimal solution set, predicting the performance of each solution under different production scenarios through logistics simulation, evaluating the impact of changes in key parameters on the optimization objectives, and outputting a final warehouse location configuration scheme with scientific decision support. This invention solves the problems of traditional methods neglecting production line collaboration, insufficient multi-objective balancing ability, and poor dynamic adaptability, achieving global, efficient, and adaptive optimization of logistics hub warehouse locations.

Description

Technical Field

[0001] This invention belongs to the field of manufacturing logistics and warehousing optimization technology, and in particular relates to a method and system for optimizing the location of logistics hubs for collaborative production of multiple production lines. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] In the intelligent transformation of manufacturing, the configuration of storage locations at key hubs between production lines is crucial for improving material flow efficiency. Existing storage location optimization technologies face significant bottlenecks in multi-production line collaborative production scenarios, and various mainstream intelligent algorithms and application solutions have obvious shortcomings: At the level of single algorithm application, particle swarm optimization is prone to getting trapped in local optima and has weak ability to handle complex scenarios with multiple constraints, making it unable to meet the needs of large-scale warehouse location optimization; deep learning and visual recognition technologies are limited by the coverage of the training set, are prone to misjudging special types of goods that are not included in the training, and lack spatial height information support, often leading to the failure of material stacking tasks; traditional genetic algorithms have high coding complexity and consume a lot of computing resources, making it difficult to meet the real-time requirements of actual production.

[0004] Meanwhile, single algorithms lack multi-objective optimization capabilities, exhibiting isolated optimization and ignoring capacity coupling and material dependencies between production lines. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, this invention provides a method and system for optimizing the location of logistics hubs for collaborative production across multiple production lines. This method can effectively handle the complex material coordination relationships between multiple production lines and achieve comprehensive optimization of multiple objectives such as inventory cost, downtime risk, and logistics efficiency while meeting various practical constraints. It also has good dynamic adaptability and computational efficiency.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a method for optimizing the location of logistics hubs for collaborative production across multiple production lines; A method for optimizing warehouse locations in logistics hubs for collaborative production across multiple production lines includes: Construct a logistics hub warehouse location optimization model for multi-production line collaborative production; Based on the logistics hub storage location optimization model, the NSGA-II algorithm is used for multi-objective collaborative optimization. The NSGA-II algorithm uses a specific chromosome coding structure, combines heuristic initialization and historical optimal solution injection to generate the first generation population, evaluates the quality of the solution through non-dominated sorting and crowding calculation, and uses genetic operations for iterative search to finally output the Pareto optimal solution set representing the trade-off relationship between different objectives. The obtained Pareto optimal solution set is verified by simulation and sensitivity analysis. The performance of each solution under different production scenarios is predicted through logistics simulation. The impact of changes in key parameters on the optimization objective is evaluated, and the final warehouse location configuration scheme with scientific decision support is output.

[0007] As a further technical solution, a logistics hub warehouse location optimization model for multi-production line collaborative production is constructed, including: The number of logistics handover areas for each production line, the number of buffer zones between production lines, and the material storage strategy matrix are used as decision variables. A multi-objective function system is established, which includes inventory costs, downtime risk costs, and logistics handling complexity. At the same time, constraints on available area, number of storage locations, and capacity are set to obtain a logistics hub storage location optimization model for multi-production line collaborative production.

[0008] As a further technical solution, the method of combining heuristic initialization with the injection of historical optimal solutions to generate the initial population is as follows: The baseline storage space demand and range are calculated based on the production capacity of each line and the area occupied by the hub, and the initial generation of individuals is randomly generated within the range. Use similar workshop hub warehouse designs from the past as high-quality initial examples.

[0009] As a further technical solution, the evaluation of the solution quality through non-dominated sorting and crowding degree calculation includes: Find all non-dominated solutions in the population and mark them as Rank1; Remove individuals from Rank 1, and find non-dominated solutions from the remaining individuals, marking them as Rank 2; Repeat this process until all individuals have been assigned a rank; For individuals within the same Rank, calculate their crowding degree in the target space. Let the crowding degree of a single individual be CD, and the formula is:

[0010] Where M is the total number of objective functions; Let i be the function value of individual i on the m-th objective; and These are the maximum and minimum values ​​of the function within the interval.

[0011] As a further technical solution, the genetic operation specifically includes: Individual selection is conducted using a tournament selection process combined with an elite retention strategy. Gene recombination was performed using a multi-point crossover strategy. An adaptive Gaussian mutation strategy is used for gene mutation, where the amount of mutation is 1 library location when performing mutation on genes related to the number of libraries.

[0012] As a further technical solution, the method also includes dynamic optimization; specifically, real-time monitoring of production line status data, material consumption rate, and storage space occupancy rate. When monitoring data indicates that the system deviates from the preset optimization state, it automatically triggers the re-solution of the optimization model or the incremental adjustment of the existing Pareto solution set to achieve dynamic adaptive updates of the storage location configuration.

[0013] The second aspect of this invention provides a logistics hub warehouse location optimization system for collaborative production across multiple production lines.

[0014] A logistics hub warehouse location optimization system for multi-production line collaborative production includes: The model building module is configured to: build a logistics hub warehouse location optimization model for multi-production line collaborative production; The collaborative solution module is configured to: perform multi-objective collaborative optimization based on the logistics hub storage location optimization model using the NSGA-II algorithm; wherein, the NSGA-II algorithm adopts a specific chromosome coding structure, combines heuristic initialization and historical optimal solution injection to generate the first generation population, evaluates the quality of the solution through non-dominated sorting and crowding calculation, and uses genetic operations for iterative search, and finally outputs a Pareto optimal solution set representing the trade-off relationship between different objectives; The decision verification module is configured to: perform simulation verification and sensitivity analysis on the obtained Pareto optimal solution set, predict the performance of each solution under different production scenarios through logistics simulation, evaluate the impact of changes in key parameters on the optimization objective, and output the final warehouse location configuration scheme supported by scientific decision-making.

[0015] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of a logistics hub location optimization method for multi-production line collaborative production as described in the first aspect of the present invention.

[0016] The fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the logistics hub storage location optimization method for multi-production line collaborative production as described in the first aspect of the present invention.

[0017] The fifth aspect of the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps in the logistics hub warehouse location optimization method for multi-production line collaborative production described in the first aspect of the present invention.

[0018] The above one or more technical solutions have the following beneficial effects: (1) This invention integrates the independent storage requirements of each production line, the buffer configuration between production lines, and the material storage strategy into a unified optimization model through chromosome coding design. It breaks the limitations of traditional "island-style" optimization, enabling the algorithm to simultaneously consider and optimize the material flow and capacity matching relationship between all production lines, effectively resolving the systemic "bucket effect" caused by congestion in single-point buffers, and improving the overall throughput capacity and system stability of the logistics hub from a global perspective. By adopting a multi-objective evolutionary algorithm framework with NSGA-II as the core, it can generate a set of evenly distributed Pareto optimal solutions in a single optimization run. The optimal solution set can show the trade-off relationship between multiple key objectives, realizing the transformation from experience-based decision-making to data-based scientific decision-making.

[0019] (2) The phased hybrid strategy and heuristic initialization method proposed in this invention specifically address the pain points of slow convergence and high computational cost in traditional genetic algorithms for site optimization. By generating a high-quality initial population based on heuristic rules of productivity and area, and injecting historical best solutions as "elites," the blind search time in the early stage of the algorithm is significantly shortened. At the same time, adaptive mutation and other mechanisms balance global exploration and local mining, effectively avoiding premature convergence and ensuring that a high-quality, near-global optimal solution is obtained within a controllable computation time.

[0020] (3) This invention not only provides a static optimization scheme, but also establishes a dynamic optimization mechanism that is linked to real-time production data (such as order changes, line status, and material consumption rate). When changes in key parameters are detected that may affect the optimization objective, the system can quickly trigger model re-optimization or make incremental adjustments based on the existing solution set. This allows the warehouse location configuration to flexibly respond to disturbances such as dynamic adjustments to the production plan and the insertion of emergency orders, ensuring the continuous effectiveness of the optimization strategy in actual operation and the overall robustness of the system.

[0021] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0022] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0023] Figure 1 This is a flowchart of the method in the first embodiment.

[0024] Figure 2 This is a flowchart illustrating the overall framework of the first embodiment.

[0025] Figure 3 The flowchart shows the NSGA-II genetic algorithm framework of the first embodiment.

[0026] Figure 4 This is a system structure diagram of the second embodiment. Detailed Implementation

[0027] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0028] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0029] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0030] This embodiment discloses a method for optimizing warehouse locations in logistics hubs for collaborative production across multiple production lines. By constructing a dynamic optimization model that includes the coupling relationships between multiple production lines, the method employs the NSGA-II genetic algorithm, which integrates heuristic initialization and historical solution injection, to solve multiple objectives collaboratively. This generates a Pareto solution set that reflects the balance between inventory costs, downtime risks, and handling efficiency. Based on simulation verification and sensitivity analysis, the method achieves optimal solution selection and ultimately supports real-time dynamic adjustment to achieve the globally optimal warehouse location configuration.

[0031] Specifically, such as Figure 1 As shown, a method for optimizing warehouse locations in logistics hubs for collaborative production across multiple production lines includes: Step S1: Construct a logistics hub warehouse location optimization model for multi-production line collaborative production.

[0032] Define decision variables using the logistics handover area corresponding to each production line as a unit. (i=1,2,...,n, where n is the total number of production lines), where This represents the number of storage locations in the logistics handover area of ​​the i-th production line. This variable directly determines the temporary storage capacity of materials for each production line, which is fundamental to ensuring continuous production. For example, in a scenario with five SPS production lines, the specific decision variable is... , ,..., This corresponds to the number of parts waiting to be shipped in the warehouse for each of the five production lines.

[0033] For the material flow requirements between any two production lines, set up a dedicated buffer and define decision variables. , representing the number of storage locations configured in the buffer zone between line p and line q. As a key node for material transfer between lines, the number of storage locations in the buffer zone must match the material supply and demand intensity between the two lines, and so on to form a complete inter-line buffer zone storage location configuration system.

[0034] Define a two-dimensional matrix As a decision variable for material storage strategy, a is the total number of material types and b is the total number of storage areas (including the logistics handover areas of each line and all inter-line buffer zones). This represents the storage proportion of material of type x in storage location area y. This matrix clarifies the distribution rules of various materials in different storage locations.

[0035] Subsequently, a multi-objective function system oriented towards "optimal cost, lowest risk, and highest efficiency" is constructed. Specifically: Total inventory holding cost is directly related to the number of storage locations and includes the cost of the space occupied by those locations, the management costs of material storage, etc. The calculation formula is as follows:

[0036] in, Let the total inventory holding cost function be... Let $\frac{1}{2}$ be the unit inventory cost of a single storage location in the logistics handover area of ​​the $i$-th production line. This function represents the unit inventory cost of a single storage location in the buffer zone between line p and line q. By quantifying the direct costs incurred by storage location occupancy, this function achieves precise control over inventory costs. For example, it considers the number of storage locations in the handover area of ​​a production line. When increasing, A corresponding increase occurs, and vice versa.

[0037] The system downtime risk cost is determined by the downtime probability and corresponding downtime losses of each production line and inter-line buffer zone. The downtime probability is negatively correlated with the number of storage locations; when insufficient storage locations lead to material supply interruptions or accumulation, the downtime probability increases significantly. The calculation formula is as follows:

[0038] in, The system downtime risk cost function. Let be the probability of production line i being shut down due to improper storage location in the handover area. The economic loss for a single shutdown of the i-th production line is... Let be the probability of line stopping due to insufficient buffer space between line p and line q. The combined economic loss from a single production line stoppage related to this buffer zone.

[0039] The complexity of logistics handling is positively correlated with the material handling distance; the longer the distance, the greater the handling difficulty, the lower the efficiency, and the higher the complexity. A weighted sum of material flow distances is used to calculate the handling complexity, with larger distances having smaller weights. The calculation formula is as follows:

[0040] in, Let the complexity function be the logistics handling process. Let x be the delivery volume of material of type x from storage location y to demand node z. Let y be the transport distance from storage location area y to demand node z. For distance weighting coefficients, satisfying Follow The function uses an inversely proportional weighting setting, where increasing the weight leads to decreasing it. By comprehensively considering material delivery volume, handling distance, and weights, it achieves a precise assessment of logistics handling efficiency. For example, allocating a certain type of frequently delivered material to a storage area closer to the demand node can significantly reduce... value.

[0041] To ensure the feasibility and practicality of the optimization model, three types of core constraints are set based on actual production scenarios: Available area constraint: The number of storage locations is limited by the actual available floor space, including the maximum available area of ​​the logistics handover area for each production line and the maximum available area of ​​the buffer zone between lines. The constraint formula is: in, Let be the floor area of ​​a single storage location in the logistics handover area of ​​the i-th production line. Let be the maximum available area of ​​the logistics handover area for the i-th production line. Let be the area occupied by a single storage location in the buffer zone between the p-th line and the q-th line. This represents the maximum usable area of ​​the buffer zone.

[0042] Storage Location Quantity Integer Constraint: The number of storage locations is a discrete physical unit and must meet the integer requirement. The constraint formula is as follows:

[0043] in, Represents the set of positive integers, i.e. and The value of can only be 1, 2, 3, ..., to ensure that the decision variable conforms to the actual application scenario.

[0044] Capacity constraints: Warehouse space allocation must match the maximum capacity of each production line to avoid resource waste caused by insufficient space limiting capacity release or excess space. The constraint formula is: in, Let T be the maximum hourly capacity of the i-th production line, and T be the material delivery cycle. Let be the maximum storage capacity of a single storage location in the handover area of ​​the i-th production line. This represents the floor function; Let p be the maximum hourly material flow rate between the p-th line and the q-th line. This represents the maximum storage capacity of a single storage location within the buffer.

[0045] Based on the above decision variables, multi-objective functions, and constraints, the following optimization model for logistics hub storage locations in multi-production line collaborative production is constructed:

[0046] This model achieves a dynamic balance between inventory cost, downtime risk, and handling complexity in warehouse location allocation by co-optimizing multiple objective functions, while satisfying various constraints. This provides a solid model foundation for the subsequent optimization solution of the NSGA-II genetic algorithm.

[0047] Step S2: Based on the logistics hub storage location optimization model, the NSGA-II algorithm is used for multi-objective collaborative optimization. The NSGA-II algorithm uses a specific chromosome coding structure, combines heuristic initialization and historical optimal solution injection to generate the first generation population, evaluates the quality of the solution through non-dominated sorting and crowding calculation, and uses genetic operations for iterative search to finally output the Pareto optimal solution set representing the trade-off relationship between different objectives.

[0048] Step S21: Considering the multi-decision variable characteristics of multi-production line collaborative storage location optimization, a multi-dimensional chromosome coding structure integrating the number of storage locations, buffer configuration, and material storage strategy is designed to ensure a complete representation of the decision variables of the optimization model. The specific structure is as follows: chromosome=

[0049] in, to The number of warehouse locations in the logistics handover area corresponding to n production lines; (n×(n 1)) / 2 gene positions correspond to the number of storage locations configured in the buffer zone between any two production lines; the last a×b gene positions correspond to the elements of the material storage strategy matrix M.

[0050] This coding structure enables the synchronous coding of three core decision variables: the number of storage locations on each production line, the configuration of buffer zones between production lines, and the material distribution strategy, providing a foundation for collaborative optimization of multiple production lines.

[0051] Step S22: To improve the quality of the initial population and shorten the algorithm's convergence time, a dual strategy of heuristic initialization and injection of historical optimal solutions is adopted to generate the first generation population. The specific steps are as follows: In heuristic initialization, the maximum capacity of each production line is first obtained. Material delivery cycle (T), storage capacity of a single warehouse location and the maximum usable area of ​​the handover area Calculate the baseline storage space requirements for the handover area of ​​each production line. Based on the constraints of the hub's land area and storage space area, a reasonable range for the number of storage spaces in each handover area is determined. ,in To meet minimum production capacity requirements Similarly, calculate the reasonable range for the number of storage locations in the buffer zone between each line. ,in , Randomly generate the number of storage sites within the above interval. The values ​​of the material storage strategy gene are randomly generated and normalized to form the basic individuals in the first generation population, accounting for 70%-80% of the total number of individuals in the first generation population.

[0052] For injecting historical optimal solutions, firstly, historical data on hub location optimization in similar production scenarios are retrieved, and the k best-performing hub location configuration schemes are selected, i.e., historical Pareto optimal solutions. These k schemes are then encoded according to the chromosome coding structure described above to form high-quality initial individuals. By verifying the uniqueness of high-quality initial individuals and heuristic initial individuals, duplicate individuals are removed and added to the first generation population, so that the proportion of high-quality individuals is 20%-30%, ultimately forming a first generation population of size N.

[0053] Step S23: The fitness of individuals is evaluated using a combination of non-dominated ranking and crowding calculation, achieving a comprehensive quantification of solution quality. The specific process is as follows: For non-dominated sorting, we first define that for any two individuals X and Y in the population, if individual X is not inferior to individual Y in all objective functions, and has at least one objective function that is strictly superior to individual Y, then X is said to dominate Y; if individual X is not dominated by any other individual in the population, then X is a non-dominated solution (Pareto optimal solution).

[0054] Subsequently, the initial population is traversed, all non-dominated solutions are selected and marked with a rank of 1. All individuals with rank=1 are removed from the population. Non-dominated solutions are selected again from the remaining individuals and marked with a rank of 2. The above process is repeated until all individuals are assigned a unique rank value. The smaller the rank value, the higher the fitness of the individual.

[0055] Finally, by calculating the crowding degree, individuals in the same Rank are evaluated differently, retaining evenly distributed individuals, maintaining population diversity, and avoiding premature convergence of the algorithm.

[0056] For a set of individuals with the same Rank, according to each objective function The values ​​are used to sort the individuals in ascending order; The crowding degree CD(i) for each individual i is calculated using the following formula:

[0057] Where M is the total number of objective functions; Let i be the function value of individual i on the m-th objective; and These represent the maximum and minimum values ​​of the function within the interval. Under the same Rank, the larger the crowding degree CD(i), the higher the fitness; conversely, the smaller the Rank value, the higher the fitness.

[0058] Step S24 involves using a combination of genetic operations, including tournament selection, multi-point crossover, and adaptive Gaussian mutation, to recombine and update the population's genes, thereby driving the algorithm toward the optimal solution.

[0059] (1) In the tournament selection process, t individuals are randomly selected from the current population to form a candidate set. The individual with the highest fitness in the candidate set is selected as the parent individual and enters the crossover pool. The above process is repeated until the number of parent individuals in the crossover pool reaches the population size N. The elite retention strategy is used to select all non-dominated solutions with Rank=1 in the current population as elite individuals. The elite individuals are directly retained to the next generation population. If the number of elite individuals is less than N, the remaining individuals are supplemented by tournament selection. If the number of elite individuals exceeds N, the N elite individuals with the highest crowding are selected to ensure that high-quality genes are not lost.

[0060] (2) Gene recombination is achieved by using multi-point crossover based on the characteristics of different gene segments on chromosomes.

[0061] Two parent individuals, P1 and P2, are randomly selected from the crossover pool, and gene segments based on the number of repositories are analyzed. and material storage strategy gene segment Set crossover points separately: 2-3 crossover points for the storage location gene segment and 'a' crossover points for the material storage strategy gene segment. Divide the chromosomes according to the crossover points, exchange the values ​​of corresponding gene segments between the two parent individuals, and generate two offspring individuals C1 and C2.

[0062] The offspring genes of the material storage strategy gene segment are normalized, and the value of the storage location quantity gene is verified to meet the constraint range. If it exceeds the range, it is corrected to the nearest constraint boundary value.

[0063] (3) To avoid premature convergence of the algorithm, an adaptive mutation strategy is designed for different gene types.

[0064] Set mutation probability Iterate through each gene locus of the offspring individuals; For the gene of storage location quantity If the randomly generated probability value is less than If so, a mutation operation is performed, and the mutation amount is fixed at 1.

[0065] After mutation, it is necessary to verify whether the gene value is within the constraint range. If the value is outside the interval, the interval boundary value is used.

[0066] For material storage strategy genes If the randomly generated probability value is less than The variation increment is then generated based on a Gaussian distribution, and all storage strategy genes for the same material type are renormalized.

[0067] Step S24: Population iteration and convergence judgment.

[0068] The parent and offspring populations are merged to form a mixed population of size 2N. The non-dominated sorting and crowding calculation are re-executed on the mixed population, and the N individuals with the highest fitness are selected to form the new generation population. A maximum number of iterations is set. If the current number of iterations reaches the maximum number of iterations, the algorithm terminates; or when no new solutions are added to the non-dominated solution set with Rank=1 in the population for k consecutive generations, the algorithm is considered to have converged and the iteration is terminated early. After the algorithm terminates, it outputs all non-dominated solutions with Rank=1 in the population, forming the Pareto optimal solution set. This solution set covers the trade-offs between the three objectives of inventory cost, downtime risk, and logistics handling complexity, providing diversified choices for actual production decisions.

[0069] Step S3 involves performing simulation verification and sensitivity analysis on the obtained Pareto optimal solution set, predicting the performance of each solution under different production scenarios through logistics simulation, evaluating the impact of key parameter changes on the optimization objective, and outputting the final warehouse location configuration scheme with scientific decision support.

[0070] We selected logistics simulation software to build a simulation model consistent with the actual production scenario. We input the warehouse location configuration parameters of each scheme in the Pareto optimal solution set, four AMR pallet delivery strategies and production constraints to simulate the logistics flow process under different production loads and material demand fluctuation scenarios. We monitored key indicators such as JPH, downtime, and handling distance to predict the actual operating performance of each scheme.

[0071] Using the number of storage locations on each production line as the key parameter, the parameter values ​​are gradually adjusted within the constraint range to analyze their impact on total inventory cost, downtime risk cost, and logistics handling complexity, thereby identifying the core sensitive parameters and their impact patterns.

[0072] Combining simulation results and sensitivity analysis conclusions, a solution with strong adaptability and excellent anti-interference ability is selected from the Pareto optimal solution set. By comprehensively balancing multiple objectives, a final warehouse location configuration scheme that can directly guide production is output, providing scientific support for decision-making.

[0073] Furthermore, in this embodiment, a dynamic optimization mechanism is added to achieve adaptive updates of storage location configuration.

[0074] The system collects data on production line operation status (such as cycle time and downtime), material consumption rate, and real-time occupancy rate of each storage location through Internet of Things (IoT) devices. The data sampling period is set to 5-10 minutes to ensure the timeliness of information.

[0075] Preset optimization thresholds (such as storage location occupancy exceeding 85% or falling below 30%, material consumption rate fluctuation exceeding 20%, and line stoppage frequency exceeding preset values) are used to compare real-time monitoring data with these thresholds to determine if the system deviates from the optimal state. If there is no deviation, the current storage location configuration is maintained; if there is a deviation, adjustments are automatically triggered: for minor deviations, incremental adjustments are made based on the existing Pareto optimal solution set to match the current operating conditions; for severe deviations, the NSGA-II algorithm is called again to solve the optimization model and update the storage location configuration parameters.

[0076] The adjusted warehouse location configuration is synchronized to the logistics management system to guide AMR delivery and warehouse location scheduling, enabling warehouse location configuration to dynamically adapt to production changes.

[0077] Example 2 This embodiment discloses a logistics hub warehouse location optimization system for multi-production line collaborative production; like Figure 2 As shown, a logistics hub warehouse location optimization system for multi-production line collaborative production includes: The model building module is configured to: build a logistics hub warehouse location optimization model for multi-production line collaborative production; Define decision variables using the logistics handover area corresponding to each production line as a unit. (i=1,2,...,n, where n is the total number of production lines), where This represents the number of storage locations in the logistics handover area of ​​the i-th production line. This variable directly determines the temporary storage capacity of materials for each production line, which is fundamental to ensuring continuous production. For example, in a scenario with five SPS production lines, the specific decision variable is... , ,..., This corresponds to the number of parts waiting to be shipped in the warehouse for each of the five production lines.

[0078] For the material flow requirements between any two production lines, set up a dedicated buffer and define decision variables. , representing the number of storage locations configured in the buffer zone between line p and line q. As a key node for material transfer between lines, the number of storage locations in the buffer zone must match the material supply and demand intensity between the two lines, and so on to form a complete inter-line buffer zone storage location configuration system.

[0079] Define a two-dimensional matrix As a decision variable for material storage strategy, a is the total number of material types and b is the total number of storage areas (including the logistics handover areas of each line and all inter-line buffer zones). This represents the storage proportion of material of type x in storage location area y. This matrix clarifies the distribution rules of various materials in different storage locations.

[0080] Subsequently, a multi-objective function system oriented towards "optimal cost, lowest risk, and highest efficiency" is constructed. Specifically: Total inventory holding cost is directly related to the number of storage locations and includes the cost of the space occupied by those locations, the management costs of material storage, etc. The calculation formula is as follows:

[0081] in, Let the total inventory holding cost function be... Let $\frac{1}{2}$ be the unit inventory cost of a single storage location in the logistics handover area of ​​the $i$-th production line. This function represents the unit inventory cost of a single storage location in the buffer zone between line p and line q. By quantifying the direct costs incurred by storage location occupancy, this function achieves precise control over inventory costs. For example, it considers the number of storage locations in the handover area of ​​a production line. When increasing, A corresponding increase occurs, and vice versa.

[0082] The system downtime risk cost is determined by the downtime probability and corresponding downtime losses of each production line and inter-line buffer zone. The downtime probability is negatively correlated with the number of storage locations; when insufficient storage locations lead to material supply interruptions or accumulation, the downtime probability increases significantly. The calculation formula is as follows:

[0083] in, The system downtime risk cost function. Let be the probability of production line i being shut down due to improper storage location in the handover area. The economic loss for a single shutdown of the i-th production line is... Let be the probability of line stopping due to insufficient buffer space between line p and line q. The combined economic loss from a single production line stoppage related to this buffer zone.

[0084] The complexity of logistics handling is positively correlated with the material handling distance; the longer the distance, the greater the handling difficulty, the lower the efficiency, and the higher the complexity. A weighted sum of material flow distances is used to calculate the handling complexity, with larger distances having smaller weights. The calculation formula is as follows:

[0085] in, Let the complexity function be the logistics handling process. Let x be the delivery volume of material of type x from storage location y to demand node z. Let y be the transport distance from storage location area y to demand node z. For distance weighting coefficients, satisfying Follow The function uses an inversely proportional weighting setting, where increasing the weight leads to decreasing it. By comprehensively considering material delivery volume, handling distance, and weights, it achieves a precise assessment of logistics handling efficiency. For example, allocating a certain type of frequently delivered material to a storage area closer to the demand node can significantly reduce... value.

[0086] To ensure the feasibility and practicality of the optimization model, three types of core constraints are set based on actual production scenarios: Available area constraint: The number of storage locations is limited by the actual available floor space, including the maximum available area of ​​the logistics handover area for each production line and the maximum available area of ​​the buffer zone between lines. The constraint formula is: in, Let be the floor area of ​​a single storage location in the logistics handover area of ​​the i-th production line. Let be the maximum available area of ​​the logistics handover area for the i-th production line. Let be the area occupied by a single storage location in the buffer zone between the p-th line and the q-th line. This represents the maximum usable area of ​​the buffer zone.

[0087] Storage Location Quantity Integer Constraint: The number of storage locations is a discrete physical unit and must meet the integer requirement. The constraint formula is as follows:

[0088] in, Represents the set of positive integers, i.e. and The value of can only be 1, 2, 3, ..., to ensure that the decision variable conforms to the actual application scenario.

[0089] Capacity constraints: Warehouse space allocation must match the maximum capacity of each production line to avoid resource waste caused by insufficient space limiting capacity release or excess space. The constraint formula is: in, Let T be the maximum hourly capacity of the i-th production line, and T be the material delivery cycle. Let be the maximum storage capacity of a single storage location in the handover area of ​​the i-th production line. This represents the floor function; Let p be the maximum hourly material flow rate between the p-th line and the q-th line. This represents the maximum storage capacity of a single storage location within the buffer.

[0090] Based on the above decision variables, multi-objective functions, and constraints, the following optimization model for logistics hub storage locations in multi-production line collaborative production is constructed:

[0091] This model achieves a dynamic balance between inventory cost, downtime risk, and handling complexity in warehouse location allocation by co-optimizing multiple objective functions, while satisfying various constraints. This provides a solid model foundation for the subsequent optimization solution of the NSGA-II genetic algorithm.

[0092] The collaborative solution module is configured to: perform multi-objective collaborative optimization based on the logistics hub storage location optimization model using the NSGA-II algorithm; wherein, the NSGA-II algorithm adopts a specific chromosome coding structure, combines heuristic initialization and historical optimal solution injection to generate the first generation population, evaluates the quality of the solution through non-dominated sorting and crowding calculation, and uses genetic operations for iterative search, and finally outputs a Pareto optimal solution set representing the trade-off relationship between different objectives; To address the multi-decision variable characteristics of collaborative warehouse location optimization across multiple production lines, a multi-dimensional chromosome coding structure integrating the number of warehouse locations, buffer configuration, and material storage strategies is designed to ensure a complete representation of the decision variables in the optimization model. The specific structure is as follows: chromosome=

[0093] in, to The number of warehouse locations in the logistics handover area corresponding to n production lines; (n×(n 1)) / 2 gene positions correspond to the number of storage locations configured in the buffer zone between any two production lines; the last a×b gene positions correspond to the elements of the material storage strategy matrix M.

[0094] This coding structure enables the synchronous coding of three core decision variables: the number of storage locations on each production line, the configuration of buffer zones between production lines, and the material distribution strategy, providing a foundation for collaborative optimization of multiple production lines.

[0095] To improve the quality of the initial population and shorten the algorithm's convergence time, a dual strategy of heuristic initialization and injection of historical optimal solutions is adopted to generate the first generation population. The specific steps are as follows: In heuristic initialization, the maximum capacity of each production line is first obtained. Material delivery cycle (T), storage capacity of a single warehouse location and the maximum usable area of ​​the handover area Calculate the baseline storage space requirements for the handover area of ​​each production line. Based on the constraints of the hub's land area and storage space area, a reasonable range for the number of storage spaces in each handover area is determined. ,in To meet minimum production capacity requirements Similarly, calculate the reasonable range for the number of storage locations in the buffer zone between each line. ,in , Randomly generate the number of storage sites within the above interval. The values ​​of the material storage strategy gene are randomly generated and normalized to form the basic individuals in the first generation population, accounting for 70%-80% of the total number of individuals in the first generation population.

[0096] For injecting historical optimal solutions, firstly, historical data on hub location optimization in similar production scenarios are retrieved, and the k best-performing hub location configuration schemes are selected, i.e., historical Pareto optimal solutions. These k schemes are then encoded according to the chromosome coding structure described above to form high-quality initial individuals. By verifying the uniqueness of high-quality initial individuals and heuristic initial individuals, duplicate individuals are removed and added to the first generation population, so that the proportion of high-quality individuals is 20%-30%, ultimately forming a first generation population of size N.

[0097] An individual fitness assessment method combining non-dominated ranking and crowding calculation is used to achieve a comprehensive quantification of solution quality. The specific process is as follows: For non-dominated sorting, we first define that for any two individuals X and Y in the population, if individual X is not inferior to individual Y in all objective functions, and has at least one objective function that is strictly superior to individual Y, then X is said to dominate Y; if individual X is not dominated by any other individual in the population, then X is a non-dominated solution (Pareto optimal solution).

[0098] Subsequently, the initial population is traversed, all non-dominated solutions are selected and marked with a rank of 1. All individuals with rank=1 are removed from the population. Non-dominated solutions are selected again from the remaining individuals and marked with a rank of 2. The above process is repeated until all individuals are assigned a unique rank value. The smaller the rank value, the higher the fitness of the individual.

[0099] Finally, by calculating the crowding degree, individuals in the same Rank are evaluated differently, retaining evenly distributed individuals, maintaining population diversity, and avoiding premature convergence of the algorithm.

[0100] For a set of individuals with the same Rank, according to each objective function The values ​​are used to sort the individuals in ascending order; The crowding degree CD(i) for each individual i is calculated using the following formula:

[0101] Where M is the total number of objective functions; Let i be the function value of individual i on the m-th objective; and These represent the maximum and minimum values ​​of the function within the interval. Under the same Rank, the larger the crowding degree CD(i), the higher the fitness; conversely, the smaller the Rank value, the higher the fitness.

[0102] By combining genetic operations such as tournament selection, multi-point crossover, and adaptive Gaussian mutation, the recombination and updating of the population's genes are achieved, driving the algorithm to converge toward the optimal solution.

[0103] (1) In the tournament selection process, t individuals are randomly selected from the current population to form a candidate set. The individual with the highest fitness in the candidate set is selected as the parent individual and enters the crossover pool. The above process is repeated until the number of parent individuals in the crossover pool reaches the population size N. The elite retention strategy is used to select all non-dominated solutions with Rank=1 in the current population as elite individuals. The elite individuals are directly retained to the next generation population. If the number of elite individuals is less than N, the remaining individuals are supplemented by tournament selection. If the number of elite individuals exceeds N, the N elite individuals with the highest crowding are selected to ensure that high-quality genes are not lost.

[0104] (2) Gene recombination is achieved by using multi-point crossover based on the characteristics of different gene segments on chromosomes.

[0105] Two parent individuals, P1 and P2, are randomly selected from the crossover pool, and gene segments based on the number of repositories are analyzed. and material storage strategy gene segment Set crossover points separately: 2-3 crossover points for the storage location gene segment and 'a' crossover points for the material storage strategy gene segment. Divide the chromosomes according to the crossover points, exchange the values ​​of corresponding gene segments between the two parent individuals, and generate two offspring individuals C1 and C2.

[0106] The offspring genes of the material storage strategy gene segment are normalized, and the value of the storage location quantity gene is verified to meet the constraint range. If it exceeds the range, it is corrected to the nearest constraint boundary value.

[0107] (3) To avoid premature convergence of the algorithm, an adaptive mutation strategy is designed for different gene types.

[0108] Set mutation probability Iterate through each gene locus of the offspring individuals; For the gene of storage location quantity If the randomly generated probability value is less than If so, a mutation operation is performed, and the mutation amount is fixed at 1.

[0109] After mutation, it is necessary to verify whether the gene value is within the constraint range. If the value is outside the interval, the interval boundary value is used.

[0110] For material storage strategy genes If the randomly generated probability value is less than The variation increment is then generated based on a Gaussian distribution, and all storage strategy genes for the same material type are renormalized.

[0111] The parent and offspring populations are merged to form a mixed population of size 2N. The non-dominated sorting and crowding calculation are re-executed on the mixed population, and the N individuals with the highest fitness are selected to form the new generation population. A maximum number of iterations is set. If the current number of iterations reaches the maximum number of iterations, the algorithm terminates; or when no new solutions are added to the non-dominated solution set with Rank=1 in the population for k consecutive generations, the algorithm is considered to have converged and the iteration is terminated early. After the algorithm terminates, it outputs all non-dominated solutions with Rank=1 in the population, forming the Pareto optimal solution set. This solution set covers the trade-offs between the three objectives of inventory cost, downtime risk, and logistics handling complexity, providing diversified choices for actual production decisions.

[0112] The decision verification module is configured to: perform simulation verification and sensitivity analysis on the obtained Pareto optimal solution set, predict the performance of each solution under different production scenarios through logistics simulation, evaluate the impact of changes in key parameters on the optimization objective, and output the final warehouse location configuration scheme supported by scientific decision-making.

[0113] We selected logistics simulation software to build a simulation model consistent with the actual production scenario. We input the warehouse location configuration parameters of each scheme in the Pareto optimal solution set, four AMR pallet delivery strategies and production constraints to simulate the logistics flow process under different production loads and material demand fluctuation scenarios. We monitored key indicators such as JPH, downtime, and handling distance to predict the actual operating performance of each scheme.

[0114] Using the number of storage locations on each production line as the key parameter, the parameter values ​​are gradually adjusted within the constraint range to analyze their impact on total inventory cost, downtime risk cost, and logistics handling complexity, thereby identifying the core sensitive parameters and their impact patterns.

[0115] Combining simulation results and sensitivity analysis conclusions, a solution with strong adaptability and excellent anti-interference ability is selected from the Pareto optimal solution set. By comprehensively balancing multiple objectives, a final warehouse location configuration scheme that can directly guide production is output, providing scientific support for decision-making.

[0116] Furthermore, in this embodiment, a dynamic optimization mechanism is added to achieve adaptive updates of storage location configuration.

[0117] The system collects data on production line operation status (such as cycle time and downtime), material consumption rate, and real-time occupancy rate of each storage location through Internet of Things (IoT) devices. The data sampling period is set to 5-10 minutes to ensure the timeliness of information.

[0118] Preset optimization thresholds (such as storage location occupancy exceeding 85% or falling below 30%, material consumption rate fluctuation exceeding 20%, and line stoppage frequency exceeding preset values) are used to compare real-time monitoring data with these thresholds to determine if the system deviates from the optimal state. If there is no deviation, the current storage location configuration is maintained; if there is a deviation, adjustments are automatically triggered: for minor deviations, incremental adjustments are made based on the existing Pareto optimal solution set to match the current operating conditions; for severe deviations, the NSGA-II algorithm is called again to solve the optimization model and update the storage location configuration parameters.

[0119] The adjusted warehouse location configuration is synchronized to the logistics management system to guide AMR delivery and warehouse location scheduling, enabling warehouse location configuration to dynamically adapt to production changes.

[0120] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0121] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in a logistics hub location optimization method for multi-production line collaborative production as described in Example 1.

[0122] Example 4 The purpose of this embodiment is to provide an electronic device.

[0123] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the logistics hub location optimization method for multi-production line collaborative production as described in Example 1.

[0124] Example 5 Embodiment 5 of the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps in the logistics hub warehouse location optimization method for multi-production line collaborative production as described in Embodiment 1.

[0125] The steps and methods involved in the apparatuses of Embodiments 2, 3, 4, and 5 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0126] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0127] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A logistics hub location optimization method for multi-line collaborative production, characterized in that, include: Construct a logistics hub warehouse location optimization model for multi-production line collaborative production; Based on the logistics hub storage location optimization model, the NSGA-II algorithm is used for multi-objective collaborative optimization. The NSGA-II algorithm uses a specific chromosome coding structure, combines heuristic initialization and historical optimal solution injection to generate the first generation population, evaluates the quality of the solution through non-dominated sorting and crowding calculation, and uses genetic operations for iterative search to finally output the Pareto optimal solution set representing the trade-off relationship between different objectives. The obtained Pareto optimal solution set is verified by simulation and sensitivity analysis. The performance of each solution under different production scenarios is predicted through logistics simulation. The impact of changes in key parameters on the optimization objective is evaluated, and the final warehouse location configuration scheme with scientific decision support is output.

2. The logistics hub location optimization method for multi-line collaborative production of claim 1, wherein, Construct a logistics hub warehouse location optimization model for multi-production line collaborative production, including: The number of logistics handover areas for each production line, the number of buffer zones between production lines, and the material storage strategy matrix are used as decision variables. A multi-objective function system is established, which includes inventory costs, downtime risk costs, and logistics handling complexity. At the same time, constraints on available area, number of storage locations, and capacity are set to obtain a logistics hub storage location optimization model for multi-production line collaborative production. 3.The method of claim 1, wherein, The process of generating the initial population by combining heuristic initialization with the injection of historical optimal solutions is as follows: The baseline storage space demand and range are calculated based on the production capacity of each line and the area occupied by the hub, and the initial generation of individuals is randomly generated within the range. Use similar workshop hub warehouse designs from the past as high-quality initial examples. 4.The method of claim 1, wherein, The evaluation of the solution quality through non-dominated sorting and crowding calculation includes: Find all non-dominated solutions in the population and mark them as Rank1; Remove individuals from Rank 1, and find non-dominated solutions from the remaining individuals, marking them as Rank 2; Repeat this process until all individuals have been assigned a rank; For individuals within the same Rank, calculate their crowding degree in the target space. Let the crowding degree of a single individual be CD, and the formula is: where M is the total number of objective functions; is the function value of individual i on the mth objective; and is the maximum, minimum of the function in the interval.

5. The multi-line oriented collaborative production facing logistics hub location optimization method of claim 1, wherein, The genetic operation specifically refers to: Individual selection is conducted using a tournament selection process combined with an elite retention strategy. Gene recombination was performed using a multi-point crossover strategy. An adaptive Gaussian mutation strategy is used for gene mutation, where the amount of mutation is 1 library location when performing mutation on genes related to the number of libraries.

6. The multi-line oriented collaborative production facing logistics hub location optimization method of claim 1, wherein, It also includes dynamic optimization, specifically, real-time monitoring of production line status data, material consumption rate, and warehouse occupancy rate; When monitoring data indicates that the system deviates from the preset optimization state, it automatically triggers the re-solution of the optimization model or the incremental adjustment of the existing Pareto solution set to achieve dynamic adaptive updates of the storage location configuration.

7. A logistics hub location optimization system for multi-line collaborative production, characterized in that, include: The model building module is configured to: build a logistics hub warehouse location optimization model for multi-production line collaborative production; The collaborative solution module is configured to: perform multi-objective collaborative optimization based on the logistics hub storage location optimization model using the NSGA-II algorithm; wherein, the NSGA-II algorithm adopts a specific chromosome coding structure, combines heuristic initialization and historical optimal solution injection to generate the first generation population, evaluates the quality of the solution through non-dominated sorting and crowding calculation, and uses genetic operations for iterative search, and finally outputs a Pareto optimal solution set representing the trade-off relationship between different objectives; The decision verification module is configured to: perform simulation verification and sensitivity analysis on the obtained Pareto optimal solution set, predict the performance of each solution under different production scenarios through logistics simulation, evaluate the impact of changes in key parameters on the optimization objective, and output the final warehouse location configuration scheme supported by scientific decision-making.

8. A computer-readable storage medium having stored thereon a program, characterized in that, When the program is executed by the processor, it implements the steps in the logistics hub storage location optimization method for multi-production line collaborative production as described in any one of claims 1-6.

9. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized by When the processor executes the program, it implements the steps in the logistics hub storage location optimization method for multi-production line collaborative production as described in any one of claims 1-6.

10. A computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of a logistics hub location optimization method for multi-production line collaborative production as described in any one of claims 1-6.

Images