Intelligent scheduling and optimizing method and system for industrial production process

Abstract

The invention relates to the technical field of industrial production scheduling optimization, and discloses an industrial production process intelligent scheduling and optimization method and system.The industrial production process intelligent scheduling and optimization method comprises the steps that resource competition hotspots are identified by establishing a resource time two-dimensional matrix; according to the method, the resource competition relationship between the emergency order and the existing order can be accurately identified, global re-planning is replaced by local fine tuning, the optimal insertion position is determined by constructing an order interference degree evaluation model, the hotspot intensity is reduced by implementing a resource competition mitigation strategy, and optimal embedding of the emergency order is realized by adopting a progressive order insertion algorithm. The disturbance to an existing production plan is minimized, the emergency order processing efficiency and the system stability are remarkably improved, and the technical problem of intelligent scheduling of emergency orders in industrial production is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial production scheduling and optimization, and more specifically, to an industrial production process intelligent scheduling and optimization method and system. Background Art

[0002] In modern industrial production environments, production planning often faces the challenge of inserting urgent orders. Urgent orders usually have higher priorities and strict delivery requirements, and need to be processed in a timely manner without excessively interfering with existing production plans.

[0003] However, the existing industrial production scheduling system has the following main technical problems when processing emergency orders: traditional scheduling methods are difficult to accurately identify the resource competition relationship between emergency orders and existing orders, resulting in improper insertion position selection, causing competition conflicts for key production resources (such as equipment, manpower, materials, etc.), and reducing the efficiency of emergency order processing.

[0004] Existing technologies mostly use a global re-planning approach to handle emergency orders, which makes it difficult to accurately assess the impact of emergency orders on existing production plans, and can easily lead to large-scale production plan adjustments, significantly increasing system instability and scheduling costs.

[0005] In addition, there are complex dependencies and uncertainties between processes in the industrial production process. Traditional scheduling methods find it difficult to efficiently insert emergency orders while taking these constraints into account. The re-planned plan may not be executed due to process dependency conflicts.

[0006] Therefore, there is an urgent need for an intelligent scheduling and optimization method that can accurately identify resource competition, minimize interference effects, and ensure that process constraints are met, so as to achieve efficient insertion of emergency orders and smooth adjustment of production plans. Summary of the invention

[0007] The present invention provides an intelligent scheduling and optimization method and system for an industrial production process, which solves the technical problems in the related art of low efficiency of emergency order insertion, great interference with existing production plans, and difficulty in ensuring the feasibility of scheduling schemes.

[0008] The present invention provides an intelligent scheduling and optimization method for an industrial production process, comprising:

[0009] Establish a two-dimensional matrix of resources and time, analyze the resource usage of urgent orders and existing orders, and identify resource competition hotspots;

[0010] Based on the identified resource competition hotspots, an order interference evaluation model is constructed, and the optimal insertion position of emergency orders is determined by combining critical path analysis and process dependency graphs.

[0011] For the identified resource contention hotspots, select and implement the optimal mitigation strategy combination from alternative resource allocation, process splitting and process route switching strategies;

[0012] According to the determined optimal insertion position and the implemented mitigation strategy combination, the urgent order processes are prioritized, an incremental process adjustment algorithm is used to achieve key process insertion, and a constraint propagation algorithm is used to ensure that the dependencies between processes are met.

[0013] In a preferred embodiment, the step of establishing a two-dimensional matrix of resource time includes:

[0014] Divide the total time span of the production plan into multiple time periods according to a predetermined granularity;

[0015] Gather all available resources in the production system, including processing equipment, human resources and materials;

[0016] Establish a two-dimensional resource-time matrix, where the matrix elements represent the degree of competition for a specific resource in a specific time period;

[0017] According to the process information of the existing order, calculate and update the value of the corresponding element in the matrix.

[0018] In a preferred embodiment, the step of identifying a set of resource contention hot spots includes:

[0019] Calculate the occupation of various resources in different time periods by the process set required for urgent orders;

[0020] Superimpose the resource requirements of urgent orders with the resource usage of existing orders and update the element values ​​in the resource time matrix;

[0021] A contention hotspot threshold is set. When the contention of a resource in a specific time period exceeds the threshold, the corresponding resource, time period, and contention are added to the contention hotspot set.

[0022] In a preferred embodiment, the step of constructing an order interference evaluation model includes:

[0023] Calculate the resource competition intensity between urgent orders and existing orders;

[0024] Calculate the degree of overlap between urgent orders and existing orders in the time dimension;

[0025] Analyze the process dependency complexity between urgent orders and existing orders;

[0026] Taking the above factors into consideration, an order interference evaluation function is constructed to quantify the impact of urgent orders on the existing order set.

[0027] In a preferred embodiment, the step of determining the optimal insertion position of the emergency order includes: representing the existing production plan as a directed graph, in which vertices represent processes and edges represent dependencies between processes; calculating the critical path by the longest path algorithm of the graph; giving priority to processes on non-critical paths as insertion points to generate a set of candidate insertion positions; verifying the feasibility of each insertion position in combination with the process dependency graph; and selecting the position with the least interference from all feasible insertion positions as the optimal insertion position.

[0028] In a preferred embodiment, the step of selecting and implementing the optimal mitigation strategy combination includes:

[0029] Build a resource contention mitigation strategy library, including alternative resource allocation strategies, process splitting strategies, and process route switching strategies;

[0030] Define a multi-objective optimization function that considers the mitigation strategy’s impact on the intensity of competitive hotspots, implementation costs, and overall production plan stability;

[0031] Use a multi-objective optimization algorithm to select the optimal combination of mitigation strategies; adjust the corresponding process schedule and resource allocation based on the selected mitigation strategy combination to reduce the intensity of competition hotspots.

[0032] In a preferred embodiment, the multi-objective optimization algorithm is an improved non-dominated sorting genetic algorithm, comprising the following steps:

[0033] Create an initial population containing multiple randomly generated strategy combinations;

[0034] Calculate the objective function value for each strategy combination;

[0035] According to the Pareto dominance relationship, the population is sorted by non-dominance and divided into different non-dominated layers;

[0036] Calculate the crowding degree of each solution in the same non-dominated layer;

[0037] Based on the non-dominated sorting level and crowding degree, select the best strategy combination to enter the next generation;

[0038] Generate new strategy combinations through crossover and mutation operations;

[0039] Repeat the above steps until the termination condition is reached, and select the optimal strategy combination from the final Pareto frontier.

[0040] In a preferred embodiment, the step of prioritizing the emergency order processes includes:

[0041] According to the criticality of the process, the process set of the urgent order is divided into a high-priority process set, a medium-priority process set and a low-priority process set;

[0042] Prioritization uses a multi-attribute decision-making method that takes into account the position of the process on the critical path, resource usage, dependencies with other processes, and the degree of impact on product quality;

[0043] A weighted scoring model is used to calculate the priority score of each process, and the processes are divided into different priority categories based on the score.

[0044] In a preferred embodiment, the step of ensuring that dependencies between processes are satisfied by using a constraint propagation algorithm comprises:

[0045] Construct a process constraint network, where the vertex set includes all processes, and the constraint set includes the forward and backward dependencies between processes and resource usage constraints;

[0046] Use fuzzy temporal reasoning method to calculate the feasible time window of the priority process;

[0047] Select the best schedule for the process within a certain time window based on resource availability and constraint satisfaction;

[0048] The constraint propagation algorithm is used to ensure that the arrangement of all processes satisfies the dependency constraints between processes. If constraint conflicts are found, a backtracking strategy is used to make adjustments.

[0049] In a preferred embodiment, an industrial production process intelligent scheduling and optimization system is used to execute the above-mentioned industrial production process intelligent scheduling and optimization method, including:

[0050] The competition hotspot identification module is used to establish a two-dimensional matrix of resource time, analyze the resource usage of urgent orders and existing orders, and identify the resource competition hotspot set;

[0051] The interference assessment module is used to build an order interference assessment model and determine the optimal insertion position of urgent orders based on critical path analysis and process dependency graphs;

[0052] Resource contention mitigation module, used to select and implement the optimal mitigation strategy combination from alternative resource allocation, process splitting and process route switching strategies;

[0053] The progressive order insertion module is used to prioritize urgent order processes, use an incremental process adjustment algorithm to insert key processes, and ensure that dependencies between processes are met through a constraint propagation algorithm.

[0054] The beneficial effects of the present invention are: by innovatively combining the technical means of competition hotspot identification, interference assessment, resource competition mitigation and progressive order insertion, the technical problems of emergency order scheduling and optimization in the industrial production process are solved, and the following technical effects are achieved:

[0055] The processing efficiency of urgent orders is improved. The competition hotspot identification algorithm and interference evaluation model of the present invention can accurately locate the resource competition relationship between urgent orders and existing orders, and provide decision support for the efficient insertion of urgent orders.

[0056] The adopted resource contention mitigation algorithm and progressive order insertion algorithm can minimize the disturbance to the existing production plan and realize local fine-tuning instead of global replanning.

[0057] The constructed process constraint network and fuzzy temporal reasoning method enable the system to accurately handle complex process dependencies and ensure the feasibility of the generated scheduling scheme through constraint propagation.

[0058] The multi-objective optimization strategy selection algorithm is able to adaptively select the optimal combination of mitigation strategies, balancing the reduction of competition hotspot intensity, implementation costs, and the impact on production plan stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of an intelligent scheduling and optimization method for an industrial production process in the present invention. DETAILED DESCRIPTION

[0060] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0061] At least one embodiment of the present invention discloses an intelligent scheduling and optimization method for industrial production processes, such as Figure 1 As shown, the following steps are included:

[0062] Step 1: Establish a two-dimensional resource-time matrix, analyze the resource usage of urgent orders and existing orders, and identify the resource competition hotspot set;

[0063] The specific implementation is as follows:

[0064] Sub-step 1.1: Establish a two-dimensional resource-time matrix model;

[0065] First, establish a two-dimensional matrix of resource time , which is used to represent the degree of competition for various production resources in different time periods. Each element in the matrix Indicates the time period Resources The degree of competition is calculated as follows:

[0066]

[0067] in, Representation Resources In time period The competitiveness of Indicates order About Resources In time period The occupancy rate, The total number of orders in the current system.

[0068] The resource-time two-dimensional matrix model is implemented as follows:

[0069] Time granularity division: divide the total time span of the production plan into a series of time periods according to the appropriate granularity (such as hours, half days or days) ,in , , Represent the 1st, 2nd, and mth time periods respectively, is the total number of time periods.

[0070] Resource collection construction: collect all available resources in the production system , including processing equipment, human resources, materials, etc. , , Respectively represent the 1st, 2nd, and nth available resources, The total number of resources.

[0071] Matrix initialization: Create a The zero matrix of is the number of resources, is the number of time periods.

[0072] Resource usage calculation: For each existing order and the processes involved , get its resource requirements , Plan execution time period and resource usage , and update the corresponding elements in the matrix:

[0073]

[0074] in, Representation Resources In time period The competitiveness of Indicates the process The amount of resources occupied, and Respectively represent the start and end time of the process.

[0075] Matrix normalization: Normalize each element in the matrix so that its value falls within the interval [0,1] to facilitate subsequent determination of competitive hotspots:

[0076]

[0077] in, represents the normalized competitiveness value, Representation Resources The maximum capacity of Representation Resources In time period degree of competitiveness.

[0078] In practical applications, the model can be appropriately adjusted according to the characteristics of the production system, such as considering factors such as resource substitutability and process priority, so as to more accurately reflect the resource competition situation.

[0079] Sub-step 1.2: Calculate the resource usage of urgent orders and existing orders;

[0080] For urgent orders , analyze the set of processes required , calculate each process Type of resources required , Estimated execution time and resource usage , forming a resource demand matrix for emergency orders :

[0081]

[0082] in, Indicates the process number. Indicates the resource type. and Respectively represent the start and end time of the process, Indicates the resource usage. Indicates the total number of operations included in the rush order.

[0083] At the same time, the existing order set is extracted from the production planning database The resource occupancy information of the existing orders is formed into a resource occupancy matrix :

[0084]

[0085] in, Indicates the order number, Indicates order The number of processes involved, Indicates the process number. Indicates the resource type. and Respectively represent the start and end time of the process, Indicates the resource usage. Indicates the total number of existing orders.

[0086] Sub-step 1.3: Resource load analysis and hotspot identification;

[0087] The resource requirement matrix for urgent orders Resource occupancy matrix with existing orders Overlay and update the resource time matrix The element values ​​in:

[0088]

[0089] in, Indicates that after adding the emergency order, resources In time period The competitiveness of Indicates the process in the rush order. Indicates the resource type used by the process. and Respectively represent the start and end time of the process. Indicates the resource usage of this process.

[0090] Setting the contention hotspot threshold ,when When Add to competitive hotspot collection middle:

[0091]

[0092] in, Indicates the threshold for determining contention hotspots, Indicates the set of identified contention hotspots, including resource types , Time period and competitiveness Three elements.

[0093] Competition Hotspot Collection The key points that may lead to resource competition after the insertion of emergency orders are accurately identified, providing precise targets for subsequent interference assessment and scheduling optimization.

[0094] Step 2: Based on the identified resource competition hotspots, an order interference evaluation model is constructed, and the optimal insertion position of the emergency order is determined by combining the critical path analysis and the process dependency graph;

[0095] This step aims to evaluate the degree of interference of the emergency order on the existing production plan and find the insertion position with the least impact. The specific implementation is as follows:

[0096] Sub-step 2.1: Construct an order interference evaluation model;

[0097] Constructing order interference evaluation function , used to quantify urgent orders For existing order collection The interference assessment model considers the following key factors:

[0098] Resource contention intensity: Based on the set of contention hotspots identified in step 1 calculate;

[0099] Time overlap: the degree of overlap between urgent orders and existing orders in the time dimension;

[0100] Process dependency complexity: the complexity of the dependencies between processes;

[0101] The interference evaluation function is defined as follows:

[0102]

[0103] in, , and Represent the first, second, and third weight coefficients respectively, satisfying ; Indicates rush orders and orders The time overlap of Indicates rush orders and orders The process between them depends on the complexity, Indicates the total number of existing orders. represents the set of competitive hot spots, Representation Resources In time period degree of competitiveness.

[0104] Sub-step 2.2: Identify potential insertion locations based on critical path analysis;

[0105] Using the critical path analysis method, potential insertion locations are identified from the existing production plan. First, the existing production plan is represented as a directed graph , where the vertex set represents the process, edge set Indicates the dependency between processes.

[0106] Compute the graph through the longest path algorithm of the graph The critical path :

[0107]

[0108] in, represents the process on the critical path, Indicates the number of processes on the critical path.

[0109] To avoid excessive impact on the processes on the critical path, the processes on the non-critical path are given priority as insertion points. Define the candidate set of insertion positions as follows:

[0110] in, Indicates the process, Indicates the insertion time point, and Respectively represent the process The earliest start time and the latest start time represent the set of processes on the non-critical path.

[0111] Sub-step 2.3: Verify the feasibility of the insertion location;

[0112] For each candidate insertion position , combined with the process dependency graph to verify its feasibility:

[0113] Build process dependency graph ,in Contains all processes of existing orders and all processes of urgent orders, Indicates the dependencies between processes.

[0114] For the insertion position , add the urgent order process to the graph according to the planned insertion time In the new dependency graph .

[0115] Check the diagram Is there a loop? If there is a loop, it means there is a circular dependency and the insertion position is not feasible; otherwise, calculate the interference degree after insertion .

[0116] The implementation of the process dependency graph model is as follows:

[0117] Node construction: All processes are taken as nodes of the graph. Each node contains attribute information such as process ID, order, execution time, resource requirements, etc.

[0118] Edge relationship definition: Create directed edges based on the dependencies between processes. There are three main types of dependencies:

[0119] Process sequence dependency: For example, process A must be completed before process B;

[0120] Resource dependency: if two processes need to use the same resource, they cannot be executed at the same time;

[0121] Material dependency: For example, the output of process A is the input of process B;

[0122] Graph structure implementation: Use adjacency lists or adjacency matrices to store graph structures. For large-scale production systems, adjacency lists are preferred to save storage space.

[0123] Ring detection algorithm: Use depth-first search (DFS) or topological sorting algorithm to detect whether there is a ring in the graph. The specific implementation includes the following steps:

[0124] First, create two collections: the visited node collection and the node collection on the current recursive path;

[0125] Traverse each node in the graph, and if the node has not been visited, call the loop detection auxiliary function on it;

[0126] In the loop detection auxiliary function, mark the current node as visited and add it to the recursive path set;

[0127] Traverse all neighbors of the current node. If the neighbor has not been visited, recursively call the loop detection auxiliary function; if the neighbor is already in the recursive path set, it means that a loop is detected;

[0128] After checking all neighbors of the current node, remove the current node from the recursive path set;

[0129] If a loop is detected during the processing of any node, the result that the loop exists is returned; otherwise, the result that the loop does not exist is returned;

[0130] Graph update mechanism: When a new process is inserted or an existing process is adjusted, the graph structure is dynamically updated, including adding new nodes and edges while maintaining the consistency of dependencies.

[0131] Through the process dependency graph model, the system can accurately verify the feasibility of the insertion location, avoid the problem of unexecutable scheduling plans due to process dependency conflicts, and improve the success rate of urgent order insertion.

[0132] Finally, from all possible insertion positions, The function selects the position with the least interference as the optimal insertion position :

[0133]

[0134] in, represents the set of all feasible insertion positions, Indicates at location The degree of disruption after inserting an emergency order.

[0135] Optimal insertion position The locations where the insertion of emergency orders has the least impact on the existing production plan are determined, providing clear targets for subsequent resource competition mitigation and order insertion.

[0136] Step 3: For the identified resource contention hotspots, select and implement the optimal mitigation strategy combination from alternative resource allocation, process splitting, and process route switching strategies;

[0137] This step aims to use intelligent resource competition mitigation algorithms to optimize resource allocation strategies and reduce the intensity of competition hotspots for the identified competition hotspots, thereby creating conditions for the smooth insertion of urgent orders. The specific implementation is as follows:

[0138] Sub-step 3.1: Construction of resource contention mitigation strategy library;

[0139] First, build a resource competition mitigation strategy library , which includes three main strategies:

[0140] Alternative resource allocation strategies : Transferring some processes to alternative resources for execution, defined as:

[0141]

[0142] in, Indicates the process, Indicates resources in the competition hotspot, Indicates an alternative resource, Represents a collection of competitive hotspot resources. Representation Resources A collection of alternative resources.

[0143] Process splitting strategy : Divide the separable process into multiple sub-processes and execute them in parallel, defined as:

[0144]

[0145] in, Indicates the original process, Represents the sub-process set after splitting, represents a set of separable processes, Indicates the number of sub-processes after splitting.

[0146] Process route switching strategy : Select an alternative routing for part of the order, defined as:

[0147]

[0148] in, Indicates the affected orders, Indicates the original process route, Indicates an alternative process route, Represents the collection of affected orders.

[0149] Sub-step 3.2: Competition hotspot mitigation strategy selection algorithm;

[0150] For each hotspot identified in step 1 , using a multi-objective optimization algorithm to select the optimal combination of mitigation strategies:

[0151] Define the optimization objective function:

[0152]

[0153] in, Represents mitigation strategy The degree of reduction in the intensity of competitive hotspots, Indicates implementation strategy The cost, Representation strategy Impact on the stability of the overall production plan, Represents a combination of mitigation strategies.

[0154] Solve multi-objective optimization problems using the Pareto optimal solution method:

[0155]

[0156] in, represents the set of feasible strategies, and are the fourth and fifth weight coefficients respectively, represents the optimal strategy combination.

[0157] The implementation of the multi-objective optimization algorithm is as follows:

[0158] Strategy encoding: Encode each possible mitigation strategy combination into a solution vector, which contains information such as strategy type, affected process, and alternative resources.

[0159] Objective function definition:

[0160] : Calculation implementation strategy Reduction in post-competition hotspot intensity:

[0161]

[0162] in, Indicates application policy Post-resources In time period degree of competitiveness.

[0163] : Calculation implementation strategy Cost:

[0164]

[0165] in, Represents a single policy The cost includes resource switching cost, process splitting cost, etc. Represents a strategy combination A single policy in .

[0166] :Calculation strategy Impact on the stability of the overall production plan:

[0167]

[0168] in, Representation strategy Process The degree of disturbance, Represents the set of orders affected, Indicates order The process in .

[0169] Solution algorithm: The improved non-dominated sorting genetic algorithm (NSGA-II) is used to solve the multi-objective optimization problem. The steps are as follows:

[0170] Initialization: Create an initial population ,Include randomly generated strategy combinations, where Indicates the population size.

[0171] Evaluation: Calculate each strategy combination The objective function value of .

[0172] Non-dominated sorting: Sort the population according to the Pareto dominance relationship and divide the population into different non-dominated layers.

[0173] Crowding calculation: In the same non-dominated layer, the crowding of each solution is calculated to maintain the diversity of the population.

[0174] Selection: Based on the non-dominated sorting level and crowding degree, select the best strategy combinations to enter the next generation.

[0175] Crossover and mutation: Generate new strategy combinations through crossover and mutation operations.

[0176] The steps from evaluation to crossover and mutation are repeated until a termination condition (such as a maximum number of iterations or a convergence criterion) is reached.

[0177] Decision making: From the final Pareto frontier, the decision maker's preferences (through weight coefficients and represents) select the optimal strategy combination .

[0178] In practical applications, algorithm parameters such as population size, crossover probability, mutation probability, etc. can be adjusted according to the characteristics of the specific problem to balance the algorithm's exploration ability and convergence speed.

[0179] Sub-step 3.3: Implementation of resource optimization and effect evaluation;

[0180] Depending on the selected combination of mitigation strategies , and adjust the production plan accordingly:

[0181] If the policy includes alternative resource allocations , then the process From the resources Migrate to Resources Execute on and update the resource time matrix accordingly The element values ​​in:

[0182]

[0183] in, Indicates the process The execution time period, Indicates the process The resource usage of Indicates resources before adjustment In time period The competitiveness of Indicates resources before adjustment In time period degree of competitiveness.

[0184] If the strategy includes process splitting , then the process Split into sub-processes , and updates the corresponding resource usage and dependencies.

[0185] If the strategy includes process route switching , then the order The process route from Switch to , and recalculate the resource usage of related processes.

[0186] After implementing the strategy, recalculate the set of competitive hot spots , and evaluate the mitigation effect:

[0187]

[0188] in, and Respectively represent the number of contention hotspots before and after mitigation, Indicates the percentage of mitigation effect.

[0189] After resource contention mitigation, the new resource allocation scheme It effectively reduces the intensity of resource competition hotspots and creates conditions for the smooth insertion of urgent orders.

[0190] Step 4: Based on the determined optimal insertion position and the implemented mitigation strategy combination, the urgent order processes are prioritized, an incremental process adjustment algorithm is used to achieve key process insertion, and a constraint propagation algorithm is used to ensure that the dependencies between processes are met;

[0191] This step aims to achieve the optimal embedding of urgent orders based on the results of the first three steps, using a progressive order insertion algorithm, while ensuring the stability and feasibility of the overall production plan. The specific implementation is as follows:

[0192] Sub-step 4.1: Prioritization of urgent order processes;

[0193] First for urgent orders Prioritize the processes in order to ensure that the core production needs are met. The division method is as follows:

[0194] According to the criticality of the process, the processes of urgent orders are grouped together. Divided into three priority subsets:

[0195] High priority process set : Contains key processes that are essential to order completion;

[0196] Medium priority process set :Includes processes that have a significant impact on quality and performance but can be moderately adjusted;

[0197] Low priority process set : Includes auxiliary processes that can be flexibly arranged;

[0198] Prioritization uses a multi-attribute decision-making approach, considering the following attributes of the process:

[0199] The position of the process on the critical path;

[0200] Resource usage of the process;

[0201] The dependency of a process on other processes;

[0202] The degree of influence of the process on product quality;

[0203] Calculate the priority score for each process using a weighted scoring model :

[0204]

[0205] in, Indicates The weight of the attribute, Indicates the process In the The ratings on the attributes, Indicates the operation to be evaluated.

[0206] The operations are divided into three categories according to their priority scores: , , ,in and are the high priority and medium priority thresholds, respectively. Indicates the process to be classified.

[0207] Sub-step 4.2: Incremental process adjustment algorithm to achieve key process insertion;

[0208] Adopt incremental process adjustment algorithm to prioritize high-priority processes and insert key processes:

[0209] For high priority process sets Each process in , based on the optimal insertion position determined in step 2 And the resource allocation plan after step 3 optimization , calculate the best scheduling time :

[0210]

[0211] in, and Respectively represent the process The earliest and latest start times, Indicates at time Arrange process For orders The degree of influence Indicates the total number of existing orders, Indicates Existing orders.

[0212] Adopting an incremental adjustment method, only replanning the processes related to the competitive hotspots and keeping other processes unchanged, thereby minimizing the disturbance to the existing production plan. Define the set of processes that need to be replanned :

[0213]

[0214] in, Represents the set of processes in the existing production plan. Represents the process in the existing production plan. Represents resources, Indicates the time period, Representation Resources In time period The competitiveness of Represents a set of contention hotspots.

[0215] right The processes in the program are partially rescheduled and the improved key block adjustment algorithm is used to minimize the rescheduling cost:

[0216]

[0217] in, and Respectively represent the processes before and after rescheduling The start time of and Respectively represent the processes before and after rescheduling Allocated resources, represents the penalty function for resource changes, and is the weight coefficient, Indicates the process that needs to be replanned.

[0218] Sub-step 4.3: Constraint propagation ensures that dependencies between processes are satisfied;

[0219] After arranging the high-priority processes, gradually arrange the medium- and low-priority processes, and ensure that the dependencies between processes are met through constraint propagation:

[0220] Constructing process constraint network , where the vertex set Contains all processes, constraint sets Includes the forward and backward dependencies between processes, resource usage constraints, etc.

[0221] For medium priority process sets Each process in Based on the constraints of the scheduled processes, the feasible time window is calculated using fuzzy temporal reasoning method. :

[0222]

[0223]

[0224] in, and Respectively represent the process The set of preceding and succeeding processes, and Respectively represent the process The earliest start time and the latest end time of Indicates the previous process The end time of Indicates subsequent process The start time of the

[0225] The implementation of the fuzzy temporal reasoning method is as follows:

[0226] Time constraint representation: Use fuzzy sets to represent time constraints, each time point There is a membership , indicating the feasibility at that time point.

[0227] Fuzzy time window construction:

[0228] For each process , initialize its fuzzy time window

[0229] In the initial state, The degree of membership within the interval is 1, and the degree of membership outside the interval is 0;

[0230] Adjust the membership function according to the priority and urgency of the process;

[0231] Fuzzy dependencies between processes:

[0232] For traditional processes Must be in process The previously completed hard constraints are converted into fuzzy constraint processes The sooner it is completed, the faster the process The sooner you start, the better;

[0233] Defining the strength of fuzzy dependencies , indicating the closeness of the dependency;

[0234] Fuzzy inference rule design:

[0235] Rule 1: If the process If there are multiple predecessor processes, its start time depends on the maximum end time of all predecessor processes;

[0236] Rule 2: If the process A previous process If the end time of The start time is also uncertain;

[0237] Rule 3: The longer the execution time of a process, the greater the uncertainty of its end time;

[0238] Fuzzy time window update algorithm: This algorithm updates the fuzzy time window of each process in an iterative manner. The specific steps are as follows:

[0239] First, a change flag is initialized to true, indicating that an iterative update is required;

[0240] When the change flag is true, loop iteration processing is performed;

[0241] At the beginning of each iteration, set the changed flag to false;

[0242] Traverse each process and process the constraints between its predecessor and successor processes;

[0243] For each predecessor process, calculate the new lower bound of the start time based on the end time of the predecessor process;

[0244] If the calculated new start time is different from the current start time, update the current start time and set the change flag to true;

[0245] For each subsequent process, calculate the new upper bound of the end time based on the start time of the subsequent process;

[0246] If the calculated new end time is different from the current end time, update the current end time and set the change flag to true;

[0247] The iterative process is repeated until the change flag remains false, indicating that the time windows of all processes have reached a stable state;

[0248] During the calculation process, operators such as maximum, minimum, addition and subtraction on fuzzy sets are used to process fuzzy time values. These operations take into account the membership of the time points so that the calculation results conform to the rules of fuzzy logic.

[0249] Scheduling decision generation:

[0250] Based on the updated fuzzy time window, calculate the optimal scheduling time for each process;

[0251] Through the defuzzification process, such as the centroid method or the maximum membership method, the fuzzy time is converted into a definite time point;

[0252] Through the fuzzy temporal reasoning method, the system can flexibly handle the dependencies between processes, especially in complex scenarios with time uncertainty and priority conflicts, and provide more robust scheduling solutions.

[0253] In the determined time window, the best scheduling time is selected for the medium priority process based on resource availability and constraint satisfaction. For the low priority process, a flexible scheduling strategy is adopted to schedule it according to the remaining resources.

[0254] The constraint propagation algorithm is used to ensure that the arrangement of all processes satisfies the dependency constraints between processes. If constraint conflicts are found, a backtracking strategy is used to make adjustments.

[0255] After completing the progressive order insertion, the system generates the final optimized scheduling plan , including the scheduling plan of all processes of urgent orders and existing orders, to achieve efficient insertion of urgent orders and stable optimization of production plans;

[0256] In one embodiment of the present invention, an application example of the aforementioned intelligent scheduling and optimization method for industrial production processes is provided:

[0257] This example focuses on an application case in a large machinery parts manufacturing company;

[0258] A large-scale mechanical parts manufacturing company mainly produces various precision mechanical parts, including bearings, gears, hydraulic cylinders, etc. The company has multiple production lines involving multiple process links such as casting, machining, heat treatment, and surface treatment. The company adopts an order-based production model and often receives urgent orders in daily operations. It needs to handle these urgent orders efficiently without affecting the existing production plan;

[0259] The company received an urgent order O_urgent that needed to be completed within 5 days, requiring the production of 100 sets of precision hydraulic cylinder components. However, there are 15 orders {O_1, O_2, …, O_15} in the current production system, which occupy most of the key production resources. How to efficiently handle this urgent order has become the main challenge facing the company.

[0260] Implementation process example:

[0261] First, the system establishes a two-dimensional matrix of resource time to analyze the resource usage of urgent orders and existing orders. As shown in Table 1, the degree of competition for some key resources in different time periods is shown:

[0262] Table 1 Resource time competition matrix (partial data)

[0263]

[0264] Setting the contention hotspot threshold ,As shown in Table 2, the following competition hotspots are identified:

[0265] Table 2. Set of identified competition hotspots

[0266]

[0267] The possible insertion positions of emergency orders are evaluated for interference, and the results are shown in Table 3:

[0268] Table 3 Insertion position interference evaluation results

[0269]

[0270] Select IP-05, the position with the least interference, as the optimal insertion position;

[0271] For the identified contention hotspots, the system adopts a variety of resource contention mitigation strategies. The results are shown in Table 4:

[0272] Table 4. Implementation results of resource contention mitigation strategy

[0273]

[0274] First, the processes of urgent orders are divided into three categories according to their priorities: high, medium, and low, as shown in Table 5:

[0275] Table 5 Priority classification of urgent order processes

[0276]

[0277] Then the system uses an incremental process adjustment algorithm to partially adjust the existing production plan to insert urgent orders. The comparison of some process arrangements before and after the adjustment is shown in Table 6:

[0278] Table 6 Comparison of process arrangements before and after adjustment (partial data)

[0279]

[0280] Technical effect verification:

[0281] By applying the intelligent scheduling and optimization method of this embodiment, the enterprise has achieved certain technical effects in processing emergency orders, as shown in Table 7 and Table 8:

[0282] Table 7 Technical effect comparison data

[0283]

[0284] Table 8 Comparison of resource utilization

[0285]

[0286] It can be seen from the above data that this implementation not only improves the processing efficiency of urgent orders, but also effectively reduces the impact on existing production, while improving resource utilization and achieving optimization and improvement of the overall performance of the production system.

[0287] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. An intelligent scheduling and optimization method for industrial production processes, characterized in that: The following steps are involved: Establish a two-dimensional matrix of resources and time, analyze the resource usage of urgent orders and existing orders, and identify resource competition hotspots; Based on the identified resource competition hotspots, an order interference evaluation model is constructed, and the optimal insertion position of emergency orders is determined by combining critical path analysis and process dependency graphs. For the identified resource contention hotspots, select and implement the optimal mitigation strategy combination from alternative resource allocation, process splitting and process route switching strategies; According to the determined optimal insertion position and the implemented mitigation strategy combination, the urgent order processes are prioritized, an incremental process adjustment algorithm is used to achieve key process insertion, and a constraint propagation algorithm is used to ensure that the dependencies between processes are met.

2. The method for intelligent scheduling and optimization of industrial production processes according to claim 1 is characterized in that: The steps of establishing a two-dimensional matrix of resource time include: Divide the total time span of the production plan into multiple time periods according to a predetermined granularity; Gather all available resources in the production system, including processing equipment, human resources and materials; Establish a two-dimensional resource-time matrix, where the matrix elements represent the degree of competition for a specific resource in a specific time period; According to the process information of the existing order, calculate and update the value of the corresponding element in the matrix.

3. The method for intelligent scheduling and optimization of industrial production processes according to claim 1 is characterized in that: The step of identifying a set of resource contention hot spots comprises: Calculate the occupation of various resources in different time periods by the process set required for urgent orders; Superimpose the resource requirements of urgent orders with the resource usage of existing orders and update the element values ​​in the resource time matrix; A contention hotspot threshold is set. When the contention of a resource in a specific time period exceeds the threshold, the corresponding resource, time period, and contention are added to the contention hotspot set.

4. The method for intelligent scheduling and optimization of industrial production processes according to claim 1, characterized in that: The step of constructing the order interference evaluation model includes: Calculate the resource competition intensity between urgent orders and existing orders; Calculate the degree of overlap between urgent orders and existing orders in the time dimension; Analyze the process dependency complexity between urgent orders and existing orders; Taking the above factors into consideration, an order interference evaluation function is constructed to quantify the impact of urgent orders on the existing order set.

5. The method for intelligent scheduling and optimization of industrial production processes according to claim 1, characterized in that: The steps of determining the optimal insertion position of the urgent order include: representing the existing production plan as a directed graph, in which vertices represent processes and edges represent dependencies between processes; calculating the critical path by the longest path algorithm of the graph; giving priority to processes on non-critical paths as insertion points, and generating a candidate set of insertion positions; verifying the feasibility of each insertion position in combination with the process dependency graph; and selecting the position with the least interference from all feasible insertion positions as the optimal insertion position.

6. The method for intelligent scheduling and optimization of industrial production processes according to claim 1, characterized in that: The steps of selecting and implementing the optimal combination of mitigation strategies include: Build a resource contention mitigation strategy library, including alternative resource allocation strategies, process splitting strategies, and process route switching strategies; Define a multi-objective optimization function that considers the mitigation strategy’s impact on the intensity of competitive hotspots, implementation costs, and overall production plan stability; Use a multi-objective optimization algorithm to select the optimal combination of mitigation strategies; adjust the corresponding process schedule and resource allocation based on the selected mitigation strategy combination to reduce the intensity of competition hotspots.

7. The method for intelligent scheduling and optimization of industrial production processes according to claim 6 is characterized in that: The multi-objective optimization algorithm is an improved non-dominated sorting genetic algorithm, comprising the following steps: Create an initial population containing multiple randomly generated strategy combinations; Calculate the objective function value for each strategy combination; According to the Pareto dominance relationship, the population is sorted by non-dominance and divided into different non-dominated layers; Calculate the crowding degree of each solution in the same non-dominated layer; Based on the non-dominated sorting level and crowding degree, select the best strategy combination to enter the next generation; Generate new strategy combinations through crossover and mutation operations; Repeat the above steps until the termination condition is reached, and select the optimal strategy combination from the final Pareto frontier.

8. The method for intelligent scheduling and optimization of industrial production processes according to claim 1, characterized in that: The step of prioritizing the urgent order processes comprises: According to the criticality of the process, the process set of the urgent order is divided into a high-priority process set, a medium-priority process set and a low-priority process set; Prioritization uses a multi-attribute decision-making method that takes into account the position of the process on the critical path, resource usage, dependencies with other processes, and the degree of impact on product quality; A weighted scoring model is used to calculate the priority score of each process, and the processes are divided into different priority categories based on the score.

9. The method for intelligent scheduling and optimization of industrial production processes according to claim 1, characterized in that: The step of ensuring that the dependencies between processes are satisfied by the constraint propagation algorithm includes: Construct a process constraint network, where the vertex set includes all processes, and the constraint set includes the forward and backward dependencies between processes and resource usage constraints; Use fuzzy temporal reasoning method to calculate the feasible time window of the priority process; Select the best schedule for the process within a certain time window based on resource availability and constraint satisfaction; The constraint propagation algorithm is used to ensure that the arrangement of all processes satisfies the dependency constraints between processes. If constraint conflicts are found, a backtracking strategy is used to make adjustments.

10. An intelligent scheduling and optimization system for industrial production processes, characterized in that: A method for intelligent scheduling and optimization of industrial production processes for executing any one of claims 1 to 9, comprising: The competition hotspot identification module is used to establish a two-dimensional matrix of resource time, analyze the resource usage of urgent orders and existing orders, and identify the resource competition hotspot set; The interference assessment module is used to build an order interference assessment model and determine the optimal insertion position of urgent orders based on critical path analysis and process dependency graphs; Resource contention mitigation module, used to select and implement the optimal mitigation strategy combination from alternative resource allocation, process splitting and process route switching strategies; The progressive order insertion module is used to prioritize urgent order processes, use an incremental process adjustment algorithm to insert key processes, and ensure that dependencies between processes are met through a constraint propagation algorithm.

Images