A production scheduling visualization method based on big data

Abstract

The application relates to the technical field of production scheduling, and discloses a production scheduling visualization method based on big data. The method first collects multi-dimensional production data such as real-time equipment states, material inventories, personnel skills and order delivery dates, generates a four-dimensional dynamic constraint model through a dynamic constraint modeling algorithm, obtains an initial scheduling scheme through a mixed integer programming and improved genetic algorithm double-driving optimization engine, inputs the initial scheduling scheme into a visualization interaction module, displays the result in a drag-and-drop Gantt chart and detects resource conflicts in real time, and updates the model after the feasibility of the adjusted scheduling scheme is verified through a dynamic recalculation algorithm. The application integrates multi-dimensional data, realizes production scheduling optimization and visualization, can effectively solve production resource conflicts, improves production efficiency, reduces costs and provides an efficient solution for enterprise production management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of production scheduling, in particular to a production scheduling visualization method based on big data. BACKGROUND

[0002] In the current thriving modern manufacturing industry, the production environment is becoming increasingly complex, and enterprises are facing unprecedented challenges. Efficient production scheduling has become a key factor for enterprises to enhance competitiveness. Traditional production scheduling methods mainly rely on manual experience, which has many drawbacks. First, manual scheduling is difficult to fully and accurately consider the many complex factors in the production process. In terms of equipment, it is difficult to real-time monitor the running state, available time window and potential failure risk of the equipment, resulting in low equipment utilization, frequent idle or overuse of equipment, increasing production costs, and possibly affecting product production progress and quality.

[0003] In terms of material management, it is difficult for manual scheduling to accurately control material inventory data, and to make timely and accurate judgments on material completeness indicators and replenishment cycles, which can easily lead to material shortages or overstock problems. Material shortages can cause production lines to stall and delay order delivery, while material overstock can occupy a large amount of capital and storage space, increasing the operating costs of enterprises.

[0004] With the diversification and individualization of market demand, the requirements for order delivery period are becoming increasingly stringent. Manual scheduling is difficult to quickly respond to changes in order delivery period while ensuring product quality, which can easily lead to delayed order delivery, damaging the reputation and customer satisfaction of enterprises. In addition, traditional scheduling methods lack visualization means, making it difficult for production managers to intuitively understand the progress and resource allocation of the entire production process, which is not conducive to timely discovery and resolution of production problems.

[0005] Although some simple production management software has been introduced by some enterprises, these software has great limitations in function. They can only analyze and process single factors, and cannot integrate multi-dimensional production data such as equipment capacity, material completeness, personnel skills and order delivery period, making it difficult to build a comprehensive and accurate production constraint model. In terms of scheduling optimization algorithms, these software mostly use simple rules or traditional algorithms, which cannot find the optimal or near-optimal scheduling scheme in complex production environments. Moreover, the visualization interface of these software is not flexible and has poor interactivity, which cannot meet the needs of users to adjust scheduling and detect resource conflicts in real time.

[0006] With the rapid development of big data and artificial intelligence technology, new opportunities have been brought to the field of production scheduling. However, the related solutions on the market still fail to fully utilize the advantages of these advanced technologies to realize the intelligentization, high efficiency and visualization of production scheduling. Therefore, developing a production scheduling visualization method and system based on big data to solve the shortcomings of traditional scheduling methods and meet the growing production management needs of modern enterprises has become an important issue to be solved in this field. SUMMARY

[0007] The present application aims to provide a production scheduling visualization method based on big data to solve the problems raised in the background art.

[0008] To achieve the above-mentioned purpose, the present application provides the following technical solution: a production scheduling visualization method based on big data, the method comprising:

[0009] Collecting a multi-dimensional production data set; the multi-dimensional production data includes real-time equipment state data, material inventory data, personnel skill data and order delivery date data; the real-time equipment state data includes equipment available time window and fault warning information, and the material inventory data includes material completeness index and replenishment cycle;

[0010] Based on the multi-dimensional production data, a four-dimensional dynamic constraint model is generated by integrating equipment capacity constraints, material completeness constraints, personnel skill constraints and order delivery date constraints through a dynamic constraint modeling algorithm;

[0011] According to the four-dimensional dynamic constraint model, an initial scheduling scheme is generated by using a mixed integer programming and improved genetic algorithm dual-driven optimization engine;

[0012] The initial scheduling scheme is input into a visualization interaction module, the scheduling result is displayed through a drag-and-drop Gantt chart, and resource conflicts in real-time detection and adjustment operations are detected;

[0013] Based on the adjusted scheduling scheme, the feasibility is verified through a dynamic recalculation algorithm, and the four-dimensional dynamic constraint model is updated.

[0014] Preferably, the four-dimensional dynamic constraint model is generated by integrating equipment capacity constraints, material completeness constraints, personnel skill constraints and order delivery date constraints through a dynamic constraint modeling algorithm, comprising:

[0015] The real-time equipment state data is subjected to availability segmentation processing to generate an equipment time occupancy rate curve;

[0016] Based on the material inventory data, the completeness gap is calculated, and a material supply dynamic constraint chain is generated in combination with the replenishment cycle;

[0017] According to the personnel skill data, the order process complexity is matched, and a skill and task association matrix is constructed.

[0018] The order delivery date data is decomposed into process level time nodes, and is fused with the equipment time occupancy rate curve, the material supply dynamic constraint chain and the skill and task association matrix to generate a four-dimensional dynamic constraint model.

[0019] Preferably, the construction method of the mixed integer programming and improved genetic algorithm double-driven optimization engine comprises:

[0020] The objective function of the mixed integer programming model is initialized to minimize the total production cost, and the four-dimensional dynamic constraint model is loaded as a constraint condition;

[0021] An improved genetic algorithm is used to design a variable-length chromosome coding mechanism, and the equipment allocation, material consumption and personnel scheduling are coded into gene fragments;

[0022] The local optimal solution of the mixed integer programming model is injected into the population mutation operation of the genetic algorithm through an alternating iteration mechanism to generate a global optimization solution set;

[0023] According to the preset convergence threshold, the initial scheduling scheme is screened.

[0024] Preferably, the method for real-time conflict detection through drag-and-drop Gantt chart comprises:

[0025] The process time sequence in the scheduling scheme is parsed to generate a task dependency tree with hierarchical relationship;

[0026] When listening to the user's drag operation, the spatio-temporal coordinates and associated resource identifiers of the adjusted process are extracted;

[0027] The resource occupancy state of adjacent nodes in the task dependency tree is traversed, and the time overlap degree and resource conflict coefficient are calculated;

[0028] The conflict area is dynamically marked by color coding, and a conflict warning prompt is triggered.

[0029] Preferably, the verification method of the dynamic recalculation algorithm comprises:

[0030] The changed process parameters in the adjusted scheduling scheme are captured, the local recalculation area is extracted, the sub-problem solution space is constructed based on the four-dimensional dynamic constraint model, and the scheduling results of the unadjusted area are retained;

[0031] A heuristic backtracking algorithm is used to traverse the feasible solutions of the sub-problem, and consistency checking is performed with the global scheduling scheme. The feasible solutions that pass the consistency checking are merged into the global scheduling scheme, and the display data of the visual interaction module is updated.

[0032] Preferably, the method for generating the equipment time occupancy rate curve comprises:

[0033] predicting a maintenance time window according to the equipment failure early warning information, and dividing the equipment available period and the equipment unavailable period;

[0034] adopting a time slice round-robin algorithm to calculate the time occupancy weight of each process on the equipment;

[0035] combining historical production data to fit the equipment load balancing curve, and optimizing the time occupancy rate segmentation precision.

[0036] Preferably, the parameter optimization method of the improved genetic algorithm comprises:

[0037] according to the complexity of the four-dimensional dynamic constraint model, self-adaptively adjusting the crossover probability and the mutation probability;

[0038] designing an elite reservation strategy, and forcibly reserving the local optimal solution output by the mixed integer programming to the next generation population;

[0039] fusing the production cost, the delivery rate achievement rate and the resource utilization rate index through a dynamic fitness function, and optimizing the chromosome screening mechanism.

[0040] Preferably, the construction method of the task dependency tree comprises:

[0041] analyzing the process pre-post relationship in order data, and generating a directed acyclic graph structure;

[0042] based on the equipment allocation result, mapping the process node to the equipment time axis, and forming a space-time dependency link;

[0043] adopting a depth-first search algorithm to traverse the dependency link, and constructing a task dependency tree with a weighted relationship.

[0044] Preferably, the execution steps of the heuristic backtracking algorithm comprise:

[0045] defining the resource conflict priority of the local recalculation area, generating a backtracking search path, pruning the invalid solution space through a constraint propagation algorithm, and narrowing the backtracking range;

[0046] adopting a memoization search technology to cache the state of the traversed solution, selecting the optimal local solution according to the verification result, and triggering the global scheduling scheme update.

[0047] Preferably, the present application further comprises a production scheduling visualization system based on big data, and the system comprises:

[0048] a data acquisition module: used for acquiring a multi-dimensional production data set, wherein the multi-dimensional production data comprises equipment real-time state data, material inventory data, personnel skill data and order delivery data;

[0049] Dynamic modeling module: configured to integrate equipment capacity constraints, material set constraints, personnel skill constraints and order delivery constraints through dynamic constraint modeling algorithm to generate a four-dimensional dynamic constraint model;

[0050] Optimization engine module: used to generate an initial scheduling scheme based on the four-dimensional dynamic constraint model using a mixed integer programming and improved genetic algorithm dual-driven optimization engine;

[0051] Visual interactive module: displays scheduling results through a drag-and-drop Gantt chart and detects resource conflicts in real-time adjustment operations;

[0052] Real-time verification module: verifies the feasibility based on the adjusted scheduling scheme through dynamic recalculation algorithm and updates the four-dimensional dynamic constraint model.

[0053] Compared with the prior art, the beneficial effects of the present application are:

[0054] In terms of data collection and integration, by collecting comprehensive real-time state data of equipment, material inventory data, personnel skill data and order delivery data, multi-dimensional key information in the production process can be obtained. The deep integration of these data breaks the isolated situation of data in each link in traditional production management. Enterprises can plan equipment maintenance in advance based on real-time state data of equipment, reduce the impact of equipment failure on production, and improve equipment utilization; according to material inventory data, accurately control material set and replenishment time, avoid material shortage or overstock, and reduce inventory cost; combined with personnel skill data and order process complexity, realize accurate matching of personnel and tasks, fully exert the professional ability of employees, and improve production efficiency.

[0055] Generating a four-dimensional dynamic constraint model is one of the key innovations of the present application. The model effectively integrates equipment capacity constraints, material set constraints, personnel skill constraints and order delivery constraints, and fully reflects various limiting conditions in the production process. Compared with traditional single constraint or simple combination constraint model, it is more close to the actual production scene, and provides a solid foundation for scheduling scheme making. Enterprises can make scientific analysis based on the model when making production plans, avoid unreasonable production plans caused by ignoring some key constraints, and ensure smooth production process.

[0056] A dual-engine optimization approach, employing both mixed-integer programming and an improved genetic algorithm, significantly improves the quality and efficiency of scheduling solutions. Mixed-integer programming optimizes objective functions such as production cost while satisfying constraints. The improved genetic algorithm enhances its search capability and convergence speed through variable-length chromosome encoding, adaptive parameter adjustment, and elite retention strategies. The combined approach enables rapid finding of near-optimal initial scheduling solutions in complex production environments, saving enterprises substantial time and resources. Compared to traditional single-algorithm optimization methods, the dual-engine optimization better balances local and global searches, preventing the algorithm from getting trapped in local optima and improving the overall quality of scheduling solutions.

[0057] The introduction of the visual interactive module greatly enhances the convenience and intuitiveness of production management. By displaying scheduling results through drag-and-drop Gantt charts, production managers can clearly understand the progress and resource allocation of the entire production process. When making scheduling adjustments, the system can detect resource conflicts in real time and provide timely feedback to users through color coding and warning prompts. This visual interactive approach enables production managers to quickly identify problems and make adjustments, significantly improving the efficiency and accuracy of production decisions. Compared to traditional text or table-based scheduling displays, the visual interactive module reduces the difficulty of production management and minimizes the occurrence of human error.

[0058] The dynamic recalculation algorithm and real-time verification module ensure the feasibility and real-time performance of the scheduling scheme. After the scheduling scheme is adjusted, the dynamic recalculation algorithm can quickly verify whether the new scheme meets the requirements of the four-dimensional dynamic constraint model. By capturing changed process parameters, extracting local recalculation regions, and using a heuristic backtracking algorithm to traverse feasible solutions, both the accuracy of verification and computational efficiency are guaranteed. Simultaneously, the four-dimensional dynamic constraint model is updated in real time, enabling the model to reflect changes in the production process promptly, providing a more accurate basis for subsequent scheduling optimization. This real-time verification and update mechanism ensures that enterprises can quickly adjust their scheduling schemes when facing various changes in the production process, guaranteeing the continuity and stability of production. Attached Figure Description

[0059] Figure 1 This is a schematic diagram illustrating the working principle of the big data-based production scheduling visualization method described in this invention.

[0060] Figure 2 A flowchart illustrating the process of generating a four-dimensional dynamic constraint model;

[0061] Figure 3 A flowchart of the real-time collision detection process for a drag-and-drop Gantt chart;

[0062] Figure 4 This is a flowchart of the verification process for the dynamic recalculation algorithm. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] Please see Figures 1-4 This invention provides a technical solution: This invention relates to a production scheduling visualization method based on big data, and the specific implementation scheme is as follows:

[0065] During the production process, production data is collected from multiple data sources. Real-time equipment status data is acquired through sensors installed on the equipment, which monitor the equipment's operating status in real time, thereby obtaining equipment availability windows and fault warning information. Material inventory data can be extracted from the enterprise's inventory management system, from which material availability indicators and replenishment cycles can be obtained. Personnel skill data is collected by compiling employee skill training records and work experience. Order delivery date data is read from the order management system. Through this multi-channel data collection approach, comprehensive and accurate multi-dimensional production data is obtained, providing a solid data foundation for subsequent scheduling analysis.

[0066] Based on the collected multi-dimensional production data, a dynamic constraint modeling algorithm is used to integrate equipment capacity constraints, material availability constraints, personnel skill constraints, and order delivery date constraints. This algorithm comprehensively considers the correlation between various data points, incorporating constraints from different dimensions into a single model to generate a four-dimensional dynamic constraint model that reflects the actual production situation, providing constraints for subsequent scheduling optimization.

[0067] A dual-drive optimization engine, combining mixed-integer programming and an improved genetic algorithm, generates an initial scheduling scheme based on a four-dimensional dynamic constraint model. This optimization engine leverages the advantages of both algorithms to find a superior production scheduling scheme while satisfying various constraints, thereby improving the utilization efficiency of production resources and reducing production costs.

[0068] The generated initial scheduling plan is input into the visual interactive module, and the scheduling results are displayed through a drag-and-drop Gantt chart. While the user adjusts the schedule, this module monitors resource conflicts in real time. The Gantt chart presents the time arrangement and resource allocation of production tasks in an intuitive graphical way, allowing users to quickly understand production progress and resource usage. When users adjust tasks, the system can promptly detect and alert users to potential resource conflicts, facilitating appropriate adjustments.

[0069] Based on the adjusted scheduling scheme, a dynamic recalculation algorithm is used to verify its feasibility. If the scheme is feasible, the four-dimensional dynamic constraint model is updated according to the new scheduling results, so that the model can reflect changes in the actual production situation in a timely manner, providing a more accurate basis for subsequent scheduling optimization and adjustment.

[0070] Example 1:

[0071] In the process of generating the four-dimensional dynamic constraint model, the real-time status data of the equipment is segmented for availability to generate the equipment time occupancy curve. The specific steps are as follows: Predict the maintenance time window based on equipment fault warning information, dividing the equipment into available and unavailable periods. Assume that the equipment fault warning information is obtained through the equipment's built-in monitoring system. When the system detects that the operating parameters of a certain component of the equipment exceed the normal range, it will issue a fault warning. Based on past maintenance experience and the equipment's operating status, the maintenance time window is predicted. For example, if the temperature of a critical component of a certain piece of equipment continues to rise and approaches the warning value, and historical data analysis predicts that maintenance will be required within the next 24 hours, then these 24 hours are the predicted maintenance time window. During this period, the equipment is unavailable, and the other times are available periods.

[0072] The time-slice round-robin algorithm is used to calculate the time occupancy weight of each process on the equipment. The formula for the time-slice round-robin algorithm is: ,in Indicates the first The time occupancy weight of each process on the equipment Indicates the first The estimated time each process will take to utilize the equipment. This indicates the total number of steps that need to be performed on this equipment. This represents the total estimated time that all processes will occupy on the equipment. This formula allows us to calculate the relative weight of each process in terms of equipment usage time, thus enabling a more rational allocation of equipment resources.

[0073] By combining historical production data to fit equipment load balancing curves, the accuracy of time occupancy segmentation is optimized. Production data from the equipment over a period of time is collected, including workload and operating efficiency at different times. Using data analysis tools, curve fitting methods, such as least squares, are employed to fit the equipment load balancing curve. This curve provides a more accurate understanding of the equipment's load status at different times, allowing for optimization of time occupancy segmentation and ensuring that the generated equipment time occupancy curve better reflects actual production conditions.

[0074] Based on material inventory data, a kitting gap is calculated, and a dynamic constraint chain for material supply is generated by combining replenishment cycles. The material kitting index reflects the availability of materials required for production. Assuming the material kitting index uses... It means that when When, it indicates that the materials are fully stocked; when When this value is 0, it indicates a material shortage. The formula for calculating the kitting gap is: ,in This indicates a kitting shortage. It is based on the material replenishment cycle. This, along with the production schedule, generates a dynamic constraint chain for material supply. For example, if the replenishment cycle for a certain material is 7 days, and the production plan requires the material to be put into use on the 10th day, then the supply of the material needs to be ensured starting from the 3rd day; otherwise, it will affect the production schedule. This forms a dynamic constraint chain for material supply.

[0075] Based on personnel skill data, order process complexity is matched to construct a skill-task correlation matrix. Personnel skill data is categorized and organized according to skill type, skill level, etc., while order process complexity is assessed and categorized into different levels such as simple, medium, and complex. Let the number of personnel be... The number of process tasks in the order is , build a Skills and Task Association Matrix ,in Indicates the first Personal opinion on the first Skill matching degree for each task. When the first task... The individual is fully capable of completing the first When it comes to the skills required for a task When some skills are possessed, they are assigned based on the level of skill mastery. arrive The value between; when not available at all, This correlation matrix provides a clear understanding of the match between personnel skills and order tasks, offering a basis for personnel scheduling.

[0076] Order delivery date data is decomposed into process-level time nodes and integrated with equipment time occupancy curves, material supply dynamic constraint chains, and skill-task association matrices to generate a four-dimensional dynamic constraint model. The order delivery date is broken down according to production processes, determining the latest start and finish times for each process. For example, if an order's delivery date is day 30 and it contains 5 processes, the delivery date is decomposed into time nodes for each process based on the process sequence and production time estimates. These process-level time nodes are then integrated with the previously generated equipment time occupancy curves, material supply dynamic constraint chains, and skill-task association matrices, comprehensively considering constraints related to equipment, materials, personnel, and time to generate a four-dimensional dynamic constraint model that accurately reflects the actual production situation.

[0077] Example 2:

[0078] The construction process of the dual-driven optimization engine of mixed integer programming and improved genetic algorithm is as follows: The objective function of the initialized mixed integer programming model is to minimize the total production cost, let the total production cost be... It consists of equipment costs Material costs Personnel costs Composed of, i.e. In actual production, equipment costs include the purchase cost and maintenance cost of equipment; material costs are the expenses for purchasing materials; and personnel costs cover employee wages and benefits. By loading a four-dimensional dynamic constraint model as constraints, the total production cost is optimized while meeting constraints such as equipment capacity, material availability, personnel skills, and order delivery dates.

[0079] An improved genetic algorithm is used to design a variable-length chromosome encoding mechanism, encoding equipment allocation, material consumption, and personnel scheduling into gene segments. Assume that during the production process... Taiwan equipment Various materials and Employees. Equipment allocation is represented by a sequence of length equal to the number of production tasks, where each element represents the equipment number assigned to that task. Material consumption is represented by a matrix associated with each production task, where each element represents the type and quantity of materials consumed by each task. Personnel scheduling is also represented by a sequence, with each element corresponding to the employee number performing the task. This information is combined and encoded into chromosomes, where each chromosome represents a possible production scheduling scheme.

[0080] A local optimum solution from a mixed-integer programming model is injected into the mutation operation of a genetic algorithm through an alternating iterative mechanism to generate a globally optimal solution set. During the iteration process, the mixed-integer programming model is first used to solve the current problem, obtaining a local optimum solution. This local optimum solution is then used as a mutation source and injected into the genetic algorithm's population. The mutation operation in the genetic algorithm randomly alters certain gene segments of the chromosome, and injecting the local optimum solution from the mixed-integer programming model can guide the genetic algorithm to search for the globally optimal solution more quickly. Assume that in the... In this iteration, the local optimal solution obtained by the mixed integer programming is: ,Will Some gene fragments in the genetic algorithm replace corresponding fragments of some chromosomes in the genetic algorithm population. Then, crossover, mutation and other operations of the genetic algorithm are performed to generate a new population. The process continues to iterate and gradually generate a global optimal solution set.

[0081] The initial scheduling scheme is selected based on a preset convergence threshold. The convergence threshold is a pre-defined parameter used to determine whether the optimization process has converged. Let the convergence threshold be... After each iteration, calculate the rate of change of the objective function value for the current population. If the rate of change of the objective function value is less than [a certain value] in multiple consecutive iterations... If the optimization process is considered convergent, then a solution with higher fitness is selected from the current population as the initial scheduling solution. Fitness can be calculated comprehensively based on indicators such as production cost and delivery time achievement rate, for example, using a fitness function. ,in and These are weighting coefficients. This is the total production cost of the current plan. It is the maximum allowable production cost. This represents the current delivery time achievement rate. This represents the ideal maximum delivery time achievement rate. This can be achieved by adjusting the weighting coefficients. and Different optimization objectives can be emphasized based on actual production needs.

[0082] Example 3:

[0083] The specific implementation process of real-time conflict detection using drag-and-drop Gantt charts is as follows: First, the start and end times of each process, as well as the pre- and post-process relationships between processes, are extracted from the scheduling scheme. Taking a simple production process as an example, assume there are processes A, B, and C, where process B can only start after process A is completed, and process C can only start after process B is completed. Based on these relationships, a directed acyclic graph (DAG) structure is generated. Then, based on the equipment allocation results, process nodes are mapped to the equipment time axis, forming a spatiotemporal dependency chain. Assuming process A is assigned to equipment 1, process B to equipment 2, and process C to equipment 1, these processes are mapped to the corresponding equipment time axis in chronological order, clarifying the temporal and spatial dependencies of the processes. Finally, a depth-first search algorithm is used to traverse the dependency chain, constructing a weighted task dependency tree. During the traversal, different weights are assigned to nodes according to the degree of dependency between processes. For example, if process B has a high dependency on process A, the edge weight between them can be set larger, thus constructing a weighted task dependency tree.

[0084] When a user drags and drops an action, the system extracts the spatiotemporal coordinates and associated resource identifiers of the process being adjusted. When a user drags and drops an action on a Gantt chart, the system captures this action in real time. By obtaining the mouse position information on the Gantt chart and combining it with the Gantt chart's coordinate system, the spatiotemporal coordinates of the process being adjusted are determined. For example, the horizontal axis of the Gantt chart represents time, and the vertical axis represents resources such as equipment or personnel; the mouse position determines the specific location of the process in both time and resource dimensions. Simultaneously, the system extracts the associated resource identifiers, such as equipment number and personnel number, from the scheduling plan's data structure for subsequent resource conflict detection.

[0085] Traverse the resource occupancy status of adjacent nodes in the task dependency tree, and calculate the time overlap and resource conflict coefficient. Let the adjusted process be... Its adjacent nodes are The formula for calculating time overlap is: ,in and They are the processes The start and end times, and They are the processes The start and end times. When When, it indicates the process. and process There is overlap in time. The resource conflict coefficient depends on the process. and process The calculation is based on the resources used, assuming the process... and process All require the use of equipment If the device is in time overlap If the available resources are insufficient to simultaneously meet the needs of two processes, the resource conflict coefficient will increase accordingly. For example, equipment... The maximum available resources within a certain time period is process and process During this time period, the equipment The resource requirements are respectively and Resource conflict coefficient (when (when) hour, .

[0086] Conflict areas are dynamically marked using color coding, triggering conflict warnings. Based on calculated time overlap and resource conflict coefficients, conflict areas in the Gantt chart are color-coded. For example, when the time overlap is high and the resource conflict coefficient exceeds a certain threshold, the conflict area is marked in red; when the time overlap is low or the resource conflict coefficient is small but conflict is still possible, it is marked in yellow. Simultaneously, the system triggers conflict warnings, notifying users of resource conflicts via pop-ups or sounds, allowing them to adjust their scheduling plans promptly.

[0087] Example 4:

[0088] The verification process of the dynamic recalculation algorithm is as follows: The changed process parameters in the adjusted scheduling scheme are captured, the local recalculation region is extracted, a sub-problem solution space is constructed based on a four-dimensional dynamic constraint model, and the scheduling results of the unadjusted region are retained. When a user adjusts the scheduling scheme, the system automatically detects which process parameters have changed, such as the start time, end time, and resource allocation. Taking a simple adjustment as an example, suppose the user advances the start time of a certain process; then the upstream and downstream processes related to that process, as well as the resources involved, may be affected. Based on these changed parameters, the local recalculation region is determined. Within this region, a sub-problem solution space is constructed based on the four-dimensional dynamic constraint model, that is, determining the feasible scheduling scheme for the processes within this region under the constraints of equipment capacity, material availability, personnel skills, and order delivery time. Simultaneously, the scheduling results of the unadjusted region are retained to avoid unnecessary recalculation of the entire scheduling scheme and improve computational efficiency.

[0089] A heuristic backtracking algorithm is used to traverse feasible solutions to subproblems and perform consistency checks with the global scheduling scheme. Feasible solutions that pass the check are merged into the global scheduling scheme, and the data displayed in the visualization module is updated. The heuristic backtracking algorithm first defines the resource conflict priority of the local recomputation region. The priority of resource conflicts is determined based on factors such as the importance and scarcity of resources. For example, resource conflicts of critical equipment have a higher priority because the shutdown of critical equipment may have a significant impact on the entire production process. A backtracking search path is generated, and the invalid solution space is pruned using a constraint propagation algorithm to narrow the backtracking range. The constraint propagation algorithm eliminates solutions that clearly do not meet the constraints during the search process based on the constraints in the four-dimensional dynamic constraint model. For example, if a process requires a certain material, but the material cannot be supplied within the required time, then the solution containing the schedule for that process is an invalid solution and can be eliminated using the constraint propagation algorithm.

[0090] A memoized search technique is employed to cache the states of traversed solutions. Based on the verification results, the optimal local solution is selected, triggering an update to the global scheduling scheme. The memoized search technique records information about solutions already traversed; when encountering the same subproblem again, previously cached results can be directly used, avoiding redundant calculations. After traversing feasible solutions to the subproblem, each feasible solution undergoes a consistency check against the global scheduling scheme. This consistency check primarily examines whether the local solution conflicts with the scheduling results in the unadjusted region and whether it meets the constraints of the entire production system. If the check passes, the feasible solution is merged into the global scheduling scheme. Finally, the data displayed in the visualization module is updated, enabling the Gantt chart to reflect the latest scheduling results in real time, facilitating user viewing and further adjustments.

[0091] Example 5:

[0092] The parameter optimization process of the improved genetic algorithm is as follows: The crossover and mutation probabilities are adaptively adjusted based on the complexity of the four-dimensional dynamic constraint model. The complexity of the four-dimensional dynamic constraint model can be measured by the number of constraints, the number of variables, and the complexity of the relationships between constraints. Let the model complexity index be... ,when A higher probability indicates a more complex model and a larger search space; in this case, it is appropriate to increase the mutation probability. This increases the diversity of the genetic algorithm during the search process and avoids getting trapped in local optima. The formula for adjusting the mutation probability can be set as follows: ,in It is the initial mutation probability. It is the initial complexity metric. It is the maximum complexity metric. This represents the maximum mutation probability. Simultaneously, appropriately reducing the crossover probability is also important. This is to prevent excessive crossover from causing the premature loss of desirable genes. The formula for adjusting the crossover probability can be set as follows: ,in It is the initial crossover probability. It represents the minimum crossover probability. This adaptive adjustment method allows the genetic algorithm to better adapt to problems of varying complexity.

[0093] An elite retention strategy is designed to forcibly preserve locally optimal solutions from the mixed-integer programming (MIB) model to the next generation of the genetic algorithm (GMA) population. In each iteration, the MIB model obtains a locally optimal solution. This solution is directly copied to the next generation of the GMA population, ensuring that these superior solutions are not lost during the evolution of the GMA. This accelerates the convergence speed of the GMA and improves its optimization efficiency. For example, if in a certain iteration, the locally optimal solution obtained by the MIB model corresponds to a lower production cost and a higher delivery rate, preserving this solution in the next generation helps guide the GMA to find a better global solution more quickly.

[0094] The chromosome selection mechanism is optimized by integrating production cost, delivery time achievement rate, and resource utilization rate indicators through a dynamic fitness function. Let the production cost be... Delivery date achievement rate Resource utilization rate Dynamic fitness function It can be represented as: .in, , , These are the weighting coefficients, and These weighting coefficients are not fixed, but are dynamically adjusted according to actual production needs and the focus at different stages.

[0095] During peak market seasons, timely order delivery is crucial to a company's reputation and market share; therefore, it is appropriate to increase the weight given to on-time delivery rate. For example, Adjusted to Accordingly, reduce the weight of production cost. and resource utilization rate weight This gives chromosome individuals that prioritize on-time order delivery a greater advantage in the selection process. Conversely, during periods of high cost pressure for businesses, the weighting of production costs can be increased. If set as This guides the genetic algorithm to prioritize scheduling solutions with lower production costs.

[0096] In calculating delivery date achievement rate If the actual order completion time is The order specifies a delivery time of ,but Resource utilization rate The calculation requires comprehensive consideration of various resources, including equipment, personnel, and materials. Taking equipment resources as an example, suppose a certain piece of equipment is planned to be used for a certain period of time. The actual effective usage time is The utilization rate of the equipment is... By taking a weighted average of the utilization rates of all equipment, the overall equipment resource utilization rate can be obtained. Similarly, the utilization rates of personnel and material resources can be calculated, and finally, a comprehensive result can be obtained. This dynamic fitness function allows for more flexible optimization of chromosome selection based on actual production needs, enabling genetic algorithms to find scheduling solutions that meet real-world production requirements more quickly.

[0097] Example 6:

[0098] This embodiment details the specific implementation of a production scheduling visualization system based on big data. The system consists of a data acquisition module, a dynamic modeling module, an optimization engine module, a visualization interaction module, and a real-time verification module.

[0099] The data acquisition module is responsible for collecting multi-dimensional production data, including real-time equipment status data, material inventory data, personnel skill data, and order delivery date data. For real-time equipment status data acquisition, sensors such as temperature, pressure, and speed sensors are deployed on various production equipment to monitor operating parameters in real time. These sensors transmit the collected data to the data acquisition terminal via wired or wireless means, where the data is then processed and stored in the database. For material inventory data, it interfaces with the company's existing inventory management system, periodically retrieving the latest material inventory quantities, material availability indicators, and replenishment cycles. Personnel skill data is collected through a dedicated employee skill management system. After employees complete various skill training programs and obtain relevant certifications, their information is entered into this system, from which the data acquisition module periodically extracts data. Order delivery date data is read from the company's order management system to ensure that the latest and most accurate order delivery time requirements are obtained.

[0100] The dynamic modeling module integrates equipment capacity constraints, material availability constraints, personnel skill constraints, and order delivery date constraints through dynamic constraint modeling algorithms to generate a four-dimensional dynamic constraint model. When handling equipment capacity constraints, it analyzes the equipment's available time window and fault warning information based on real-time equipment status data. For example, if a piece of equipment is expected to require scheduled maintenance within the next week, lasting two days, its capacity will be set to unavailable during these two days when building the model. For material availability constraints, it calculates availability gaps based on material inventory data. If a product requires materials A, B, and C, and material A is currently in short supply, a availability gap will occur. This gap is then combined with replenishment cycles to determine material supply time nodes, forming a dynamic constraint chain for material supply. For personnel skill constraints, it matches order process complexity based on personnel skill data. For example, if an order's production process requires advanced welding skills, the model identifies personnel with these skills and establishes a skill-task relationship. Finally, order delivery date data is decomposed into process-level time nodes and integrated with other constraint factors to generate a four-dimensional dynamic constraint model that accurately reflects actual production constraints.

[0101] The optimization engine module generates an initial scheduling scheme based on a four-dimensional dynamic constraint model, employing a dual-drive optimization engine of mixed-integer programming and an improved genetic algorithm. First, the objective function of the mixed-integer programming model is initialized to minimize total production cost, incorporating equipment cost, material cost, and personnel cost into the calculation. For example, equipment cost includes depreciation and maintenance costs, material cost covers procurement and transportation costs, and personnel cost includes employee wages and benefits. The four-dimensional dynamic constraint model is then loaded as a constraint condition to ensure the scheduling scheme is optimized while meeting various actual production constraints. An improved genetic algorithm is used to design a variable-length chromosome encoding mechanism, encoding equipment allocation, material consumption, and personnel scheduling as gene segments. In practice, equipment allocation may use numbers 1, 2, 3, etc., to represent different equipment, and the assignment of tasks to corresponding equipment is encoded as part of a chromosome; material consumption is encoded according to the type and quantity of materials; personnel scheduling is similarly represented by a specific code indicating the correspondence between personnel and tasks. Through an alternating iterative mechanism, the local optimal solution of the mixed-integer programming model is injected into the population mutation operation of the genetic algorithm, continuously optimizing the population and selecting the initial scheduling scheme based on a preset convergence threshold.

[0102] The visualization and interaction module displays scheduling results via a drag-and-drop Gantt chart and detects resource conflicts during adjustments in real time. When displaying scheduling results, the Gantt chart presents the start and end times of each production task and the allocated resources in an intuitive graphical format. The horizontal axis represents time, and the vertical axis can be equipment, personnel, or other resource categories. When a user drags and drops a task in the Gantt chart, the system immediately activates the real-time conflict detection function. It parses the time sequence of processes in the scheduling plan to generate a hierarchical task dependency tree. For example, a production process may contain multiple processes with sequential relationships; these relationships are parsed to generate a task dependency tree. While listening for user drag-and-drop operations, the system extracts the spatiotemporal coordinates and associated resource identifiers of the adjusted process, traverses the resource occupancy status of adjacent nodes in the task dependency tree, and calculates the time overlap and resource conflict coefficient. If two processes occupy the same resources within the same time period, and the resource quantity is insufficient to meet both needs, a resource conflict is identified. Conflict areas are dynamically marked with color coding; red indicates a severe conflict, and yellow indicates a potential conflict, triggering conflict warnings to remind users to make timely adjustments.

[0103] The real-time verification module verifies the feasibility of the adjusted scheduling plan using a dynamic recalculation algorithm and updates the four-dimensional dynamic constraint model. After the user adjusts the scheduling plan, the real-time verification module captures the changed process parameters in the adjusted plan and extracts the local recalculation regions. For example, if the user adjusts the start time of a process, the upstream and downstream processes related to that process, as well as the resources involved, may be affected; these affected areas are the local recalculation regions. A sub-problem solution space is constructed based on the four-dimensional dynamic constraint model, retaining the scheduling results of unadjusted regions to improve computational efficiency. A heuristic backtracking algorithm traverses the feasible solutions of the sub-problems and performs consistency checks with the global scheduling plan. If the check passes, the feasible solutions are merged into the global scheduling plan, and the four-dimensional dynamic constraint model is updated to reflect the latest production situation, providing a more accurate basis for subsequent scheduling optimization and adjustments.

[0104] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0105] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A big data based production scheduling visualization method, characterized by, The application relates to a multi-dimensional production data collection method, and a multi-dimensional production data collection device. The multi-dimensional production data collection method comprises the following steps: collecting multi-dimensional production data; the multi-dimensional production data comprises real-time equipment state data, material inventory data, personnel skill data and order delivery date data; the real-time equipment state data comprises equipment available time window and fault early warning information, and the material inventory data comprises material completeness index and replenishment cycle; Based on the multi-dimensional production data, a four-dimensional dynamic constraint model is generated by integrating equipment capacity constraints, material completeness constraints, personnel skill constraints and order delivery date constraints through a dynamic constraint modeling algorithm; According to the four-dimensional dynamic constraint model, an initial scheduling scheme is generated by using a mixed integer programming and improved genetic algorithm dual-driven optimization engine; The initial scheduling scheme is input into a visual interactive module, scheduling results are displayed through a drag-and-drop Gantt chart, and resource conflicts in real-time adjustment operations are detected; The method for real-time conflict detection through the drag-and-drop Gantt chart comprises the following steps: A task dependency tree with hierarchical relationship is generated by analyzing the process time sequence in the scheduling scheme; When a user drag operation is listened to, the space-time coordinates and associated resource identifiers of the adjusted process are extracted; The resource occupation states of adjacent nodes in the task dependency tree are traversed, and the time overlap degree and resource conflict coefficient are calculated; Conflict areas are dynamically marked by color coding, and a conflict early warning prompt is triggered; Based on the adjusted scheduling scheme, the feasibility is verified through a dynamic recalculation algorithm, and the four-dimensional dynamic constraint model is updated; The verification method of the dynamic recalculation algorithm comprises the following steps: The changed process parameters in the adjusted scheduling scheme are captured, a local recalculation area is extracted, a sub-problem solving space is constructed based on the four-dimensional dynamic constraint model, and the scheduling results of the unadjusted area are retained; A heuristic backtracking algorithm is used to traverse the feasible solutions of the sub-problem, and consistency checking is performed with the global scheduling scheme, the feasible solutions that pass the consistency checking are merged into the global scheduling scheme, and the display data of the visual interactive module is updated.

2. The production scheduling visualization method based on big data according to claim 1, characterized in that, The method for generating the four-dimensional dynamic constraint model by integrating equipment capacity constraints, material completeness constraints, personnel skill constraints and order delivery date constraints through the dynamic constraint modeling algorithm comprises the following steps: The real-time equipment state data is subjected to availability segmentation processing, and an equipment time occupancy rate curve is generated; Based on the material inventory data, the completeness gap is calculated, and a material supply dynamic constraint chain is generated in combination with the replenishment cycle; According to the personnel skill data, the order process complexity is matched, and a skill and task correlation matrix is constructed; The order delivery date data is decomposed into process-level time nodes, and is fused with the equipment time occupancy rate curve, the material supply dynamic constraint chain and the skill and task correlation matrix to generate the four-dimensional dynamic constraint model.

3. The production scheduling visualization method based on big data according to claim 1, characterized in that, The construction method of the mixed integer programming and improved genetic algorithm dual-driven optimization engine comprises the following steps: The objective function of the mixed integer programming model is initialized as minimizing the total production cost, and the four-dimensional dynamic constraint model is loaded as a constraint condition; An improved genetic algorithm is used to design a variable-length chromosome coding mechanism, and equipment allocation, material consumption and personnel scheduling are coded into gene fragments; Through an alternative iteration mechanism, the local optimal solution of the mixed integer programming model is injected into the population mutation operation of the genetic algorithm to generate a global optimization solution set; According to a preset convergence threshold, the initial scheduling scheme is screened.

4. The production scheduling visualization method based on big data according to claim 2, characterized in that, The method for generating the equipment time occupancy curve comprises: According to the equipment failure early warning information, the maintenance time window is predicted, and the available period and the unavailable period of the equipment are divided; The time slice rotation algorithm is used to calculate the time occupancy weight of each process on the equipment; The device load balancing curve is fitted combined with historical production data to optimize the segmentation accuracy of the time occupancy rate.

5. The production scheduling visualization method based on big data according to claim 3, characterized in that, The parameter optimization method of the improved genetic algorithm comprises: According to the complexity of the four-dimensional dynamic constraint model, the crossover probability and the mutation probability are adaptively adjusted; An elite reservation strategy is designed to force the local optimal solution output by the mixed integer programming to be reserved to the next generation population; Through the dynamic fitness function, the production cost, the delivery rate and the resource utilization rate are fused to optimize the chromosome screening mechanism.

6. The production scheduling visualization method based on big data according to claim 1, wherein, The method for constructing the task dependency tree comprises: The precedence relationship of the processes in the order data is analyzed to generate a directed acyclic graph structure; Based on the equipment allocation result, the process nodes are mapped to the equipment time axis to form a space-time dependency link; The depth-first search algorithm is used to traverse the dependency link to construct a task dependency tree with weighted relationships.

7. The production scheduling visualization method based on big data according to claim 1, characterized in that, The execution steps of the heuristic backtracking algorithm comprise: The resource conflict priority of the local recalculation area is defined, the backtracking search path is generated, the constraint propagation algorithm is used to prune the invalid solution space, and the backtracking range is reduced; The memory search technology is used to cache the state of the traversed solution, the optimal local solution is selected according to the verification result, and the global scheduling scheme is updated.

8. A big data based production scheduling visualization system, characterized by, It comprises: A data acquisition module is configured to acquire a multi-dimensional production data set, wherein the multi-dimensional production data comprises real-time state data of equipment, material inventory data, personnel skill data and order delivery date data; A dynamic modeling module is configured to integrate equipment capacity constraints, material matching constraints, personnel skill constraints and order delivery date constraints by a dynamic constraint modeling algorithm to generate a four-dimensional dynamic constraint model; An optimization engine module is configured to generate an initial scheduling scheme by using a mixed integer programming and an improved genetic algorithm double-driven optimization engine according to the four-dimensional dynamic constraint model; A visual interactive module is configured to display the scheduling result by a drag-and-drop Gantt chart and to detect resource conflicts in real time during adjustment operations; The method for real-time conflict detection by a drag-and-drop Gantt chart comprises: The process time sequence in the scheduling scheme is analyzed to generate a task dependency tree with hierarchical relationships; When a user's drag-and-drop operation is listened to, the space-time coordinates and the associated resource identifiers of the adjusted process are extracted; The resource occupancy states of adjacent nodes in the task dependency tree are traversed to calculate the time overlap degree and the resource conflict coefficient; The conflict area is dynamically marked by color coding, and a conflict warning prompt is triggered; A real-time verification module is configured to verify the feasibility by a dynamic recalculation algorithm based on the adjusted scheduling scheme and to update the four-dimensional dynamic constraint model; The verification method of the dynamic recalculation algorithm comprises: The changed process parameters in the adjusted scheduling scheme are captured, the local recalculation area is extracted, the sub-problem solution space is constructed based on the four-dimensional dynamic constraint model, and the scheduling result of the unadjusted area is reserved; The heuristic backtracking algorithm is used to traverse the feasible solutions of the sub-problem, and consistency verification is performed with the global scheduling scheme, the feasible solutions that pass the verification are merged into the global scheduling scheme, and the display data of the visual interactive module is updated.

Images