Discrete manufacturing workshop efficient scheduling and simulation method based on connected graph

Abstract

The invention discloses a discrete manufacturing workshop efficient scheduling and simulation method based on a connected graph. The method specifically comprises the following steps: defining a workshop production execution logic model PELM-DaE based on directed service node pairs and encapsulated service units; a connected graph is constructed, and a process connected graph and a job connected graph are respectively constructed under the driving of a process route and a scheduling scheme based on PELM-DaE, so that a flexible connected graph of production execution logic is realized, a dynamic execution relationship and constraint between jobs are expressed, pre-calculation of FICM is realized, and efficient scheduling and simulation are supported; scheduling and simulation are carried out based on the connected graph, and the process connected graph is applied to job scheduling, so that reduction of candidate equipment in the job scheduling process is realized; the operation connected graph is applied to production simulation, so that the simulation calculation efficiency and the algorithm maintainability are improved. The invention provides a new technical scheme for improving the scheduling simulation efficiency of the discrete manufacturing workshop, and has important practical application value.

Description

Technical Field

[0001] The invention belongs to the technical field of discrete manufacturing, and in particular relates to an efficient scheduling and simulation method for discrete manufacturing workshops based on a connectivity graph. Background Art

[0002] Discrete manufacturing workshops need to continuously apply key digital technologies such as scheduling and simulation during production execution to achieve synchronization, prediction and decision-making of workshop production. However, due to dynamic uncertainty and interference in the actual production process, scheduling and simulation calculations need to be performed frequently, which poses a challenge to the computational efficiency of traditional methods.

[0003] At present, the scheduling and simulation technologies for discrete manufacturing workshops mainly include the following categories:

[0004] In terms of scheduling algorithms, a variety of solutions have been developed, including heuristic algorithms (such as priority rules), meta-heuristic algorithms (such as simulated annealing and genetic algorithms), reinforcement learning methods, and hybrid optimization algorithms, etc. These algorithms are mainly focused on optimizing search performance in terms of speed and accuracy to find approximate optimal solutions.

[0005] In terms of simulation technology, discrete event simulation is widely used in production process simulation, mainly including event-driven, activity scanning and process interaction methods. These methods can effectively verify the scheduling plan.

[0006] However, the prior art has the following technical problems:

[0007] 1. The existing scheduling algorithms lack comprehensive consideration of the production execution logic, making it difficult to accurately describe the execution relationship and constraints between jobs, which affects the efficiency and accuracy of scheduling calculations.

[0008] 2. In terms of scheduling calculations, existing algorithms are difficult to achieve fast response. Although a variety of optimization algorithms have been developed, due to the lack of accurate description of the execution relationship, the algorithm is difficult to effectively use constraint information during the search process, resulting in low computing efficiency. Especially in scenarios that require frequent rescheduling, existing algorithms are difficult to meet the requirements of real-time response.

[0009] 3. In terms of production simulation, the existing methods are not well integrated with the scheduling system. Usually, independent models are built for scheduling and simulation, and interfaces are developed to achieve interaction. There is a lack of a unified model to express job execution relationships and constraints, resulting in low simulation verification efficiency. Summary of the invention

[0010] In view of the above problems, the present invention provides an efficient scheduling and simulation method for discrete manufacturing workshops based on a connectivity graph.

[0011] The present invention provides a method for efficient scheduling and simulation of discrete manufacturing workshops based on a connectivity graph, comprising the following steps:

[0012] Step 1: Define the workshop production execution logic model PELM-DaE based on directed service node pairs and encapsulated service units, formalize the execution logic of the discrete manufacturing workshop, and realize the integrated description of the workshop information flow, control flow and material flow FICM.

[0013] Step 2: Construct a connectivity graph. Based on PELM-DaE, the process connectivity graph and the job connectivity graph are constructed respectively under the drive of the process route and the scheduling plan to realize the flexible connectivity graph of the production execution logic, express the dynamic execution relationship and constraints between jobs, and then realize the pre-calculation of FICM to support efficient scheduling and simulation.

[0014] Step 3: Perform scheduling and simulation based on the connectivity graph. By applying the process connectivity graph to job scheduling, the candidate equipment for the job scheduling process can be reduced. By applying the job connectivity graph to production simulation, the simulation calculation efficiency and algorithm maintainability can be improved.

[0015] Furthermore, step 1 is specifically as follows: by defining a directed service node pair DSNP and an encapsulated service unit ESC, and combining them to form PELM-DaE, an effective expression of the workshop production execution logic is achieved.

[0016] First, define the service node as follows:

[0017] SN= <P n ,A n ,Pos,T s ,S n ,F set ,Des in ,Des out > (1)

[0018] Among them, P n A is the service node type, which is divided into path node, processing node, storage node and auxiliary node according to the different services performed by the service node; A n is the attribute element of the service node, including logistics path, processing node, and buffer zone; Pos is the coordinate of the service node, indicating the specific coordinate position where the material arrives and stops during the production process; T s S is the service time, which means the time that the material stays in the service node and performs production activities after entering the service node; n is the status of the service node, including idle, working and blocked; F set It is the set of flow entities currently executing services in the node; in Indicates the conditions for the flow entity to enter the service node; Des outIndicates the conditions under which a flow entity can leave a service node and proceed to the next service node.

[0019] The definition of directed service node pair DSNP is as follows:

[0020] DSNP= <P p ,N 1 ,N 2 ,Dir,T l ,S e ,S p ,E> (2)

[0021] Among them, P p It is the type of DSNP, describing different production execution activities according to the type of service node and its connection relationship; N 1 and N 2 are two service nodes defined, representing two locations in the material space where production activities will occur; Dir is N 1 and N 2 The direction of the connecting edge between them indicates the flow relationship of materials between service nodes, which can be bidirectional, forward or reverse; l is the logistics time, which means the time required to perform material transfer activities between two service nodes; S e is the state of the DSNP connection edge, including idle, working and blocked; S p It is the state of DSNP, including activation and inactivation, reflecting the execution process of production activities; E is the executor of DSNP, which represents the logistics equipment that performs material transfer activities between two service nodes.

[0022] Two types of encapsulated service units ESC are defined, namely process encapsulation service unit PESC and buffer encapsulation service unit BESC, which are defined as follows:

[0023]

[0024]

[0025] Among them, C, P and B are control device, processing device and cache device respectively; It is a set of actuators of ESC, used to realize the material flow; DSNP set It is a DSNP set in ESC, which is used to describe the production execution logic of the unit; LN set It is the set of internal logistics paths of ESC, which is used to realize the self-organization of actuators on the ESC path; and It is the infeed and outfeed location of the unit, representing the material interaction interface between the ESC and its external elements.

[0026] For the process encapsulation service unit PESC, and Is its private input and output buffer; for the buffer encapsulation service unit BESC, PESC set It is a collection of PESCs for which storage services can be provided.

[0027] The production execution logic model PELM-DaE based on DSNP and ESC is defined as follows:

[0028]

[0029] Among them, PESC set and BESC set It is a collection of PESC and BESC, describing the composition of various production function modules in the workshop; It is a collection of DSNPs at the workshop level; and It is a collection of workshop-level logistics paths and actuators, which together describe the interaction between workshop ESCs and are a mapping of workshop production activities; F set Is a fluid entity set.

[0030] Furthermore, step 2 constructs the process connectivity diagram and the operation connectivity diagram as follows:

[0031] The process connectivity graph is a set of DSNP sequences, representing the set of possible production activity execution routes for a product driven by its process, i.e., possible future FICM, which is defined as follows:

[0032] PCM={P 1 ,P 2 ,P i ,…,P n} (6)

[0033]

[0034] Among them, P i Indicates the DSNP sequence set contained in each process of the product, Indicates the current DSNP set, Indicates the set of subsequent DSNPs to which the current DSNP is connected.

[0035] The operation connectivity graph is a subgraph of the process connectivity graph, which reduces the connectivity between a certain class of optional ESCs to a specific ESC; specifically, a process connectivity graph belongs to a certain class of products, and each material entity has an operation connectivity graph; its definition and construction process is similar to the process connectivity graph. Figure 1 To.

[0036] Furthermore, in step 3, the operation scheduling based on the process connectivity diagram is specifically implemented as follows:

[0037] Before scheduling operations, a process connectivity diagram is constructed to evaluate the rationality of the process route and the feasibility of production execution, thereby avoiding misjudgment of production capacity.

[0038] Subsequently, scheduling calculations are performed on the materials evaluated through process connectivity to generate job plans; the process connection diagram expresses the execution relationships and constraints between and within jobs by pre-calculating FICM, which is used as input for job scheduling and provides valuable insights.

[0039] Furthermore, in step 3, the production simulation based on the job connectivity graph is implemented as follows:

[0040] In order to test the feasibility of the scheduling results and determine whether the orders can be delivered on time according to the production plan, the production execution process is simulated; it will be driven by a discrete event simulation algorithm to advance each DSNP and simulation clock by inputting the job connection diagram of each material and simulating production activity events until all orders are completed.

[0041] After the simulation is completed, if the predicted order cannot be completed on time, it is necessary to return to the scheduling calculation step to readjust the production plan; the verified scheduling results and the implementation path of the job connectivity graph will be input into the actual workshop as FICM's prediction to drive production execution.

[0042] The beneficial technical effects of the present invention are:

[0043] The present invention provides a new method for constructing a connectivity graph to characterize job execution relationships and constraints, and a new framework for realizing efficient scheduling and simulation by dynamically constructing and continuously applying the connectivity graph. The present invention provides a new technical solution for improving the efficiency of discrete manufacturing workshop scheduling simulation, and has important practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a block diagram for realizing the efficient scheduling and simulation method of discrete manufacturing workshops based on connectivity graph.

[0045] Figure 2 The overall architecture of the PELM-DaE production execution logic model.

[0046] Figure 3 This is an example diagram of DSNP.

[0047] Figure 4 It is a demonstration line for intelligent manufacturing.

[0048] Figure 5 IMDL graphical representation and construction case of PELM-DaE (a is a screenshot of the process connectivity evaluation interface, b is a screenshot of the 3D PELM-DaE modeling interface, and c is a screenshot of the 2D PELM-DaE model mapping interface).

[0049] Figure 6 Generate cases for process connectivity diagrams used for job scheduling (where a is the IMDL production execution logic description based on PELM-DaE, b is the process connectivity diagram of product 5, and c is the process connectivity diagram of product 7).

[0050] Figure 7 Generate cases for job connectivity graphs for production simulation. DETAILED DESCRIPTION

[0051] The present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0052] The present invention provides a method for efficiently scheduling and simulating discrete manufacturing workshops based on a connectivity graph. Figure 1 As shown in the figure, a PLEM-DaE model is proposed to express the production execution logic in discrete manufacturing workshops, further promoting the effective description of the workshop production execution process. The specific steps are:

[0053] Step 1: Build a workshop production scheduling and simulation framework based on connectivity graph. Production execution starts with the order placement, and the production plan is arranged according to the process route of each material product in the current order, and then verified and sent to the workshop for execution. The core links of this process are data preparation, job scheduling and simulation verification. This paper proposes a scheduling and simulation framework based on connectivity graph to solve these three problems, such as Figure 2 The key technologies are: building a workshop production execution logic model, building a connectivity graph, job scheduling and production simulation.

[0054] Step 2: Construct the workshop production execution logic model. The present invention realizes the effective expression of workshop production execution logic by defining directed service node pairs (DSNP) and encapsulated service cells (ESC) and combining them to form a workshop production execution logic model (Production Execution Logic Model with Directedservice node pairs and Encapsulated service cells, PELM-DaE).

[0055] First, define the service node, which is an important part of DSNP. The definition is as follows:

[0056] SN= <P n ,A n ,Pos,T s ,S n ,Fset ,Des in ,Des out > (1)

[0057] Among them, P n It is the service node type, which is divided into path node, processing node, storage node and auxiliary node according to the different services performed by the service node.

[0058] A n It is the attribute elements of the service node, including logistics path, processing node and buffer zone.

[0059] Pos is the coordinate of the service node, indicating the specific coordinate position where the material arrives and stops during the production process.

[0060] T s Service time refers to the time that materials stay in the service node and perform production activities after entering the service node.

[0061] S n It is the status of the service node, including idle, working and blocked.

[0062] F set It is the set of stream entities currently executing services in the node.

[0063] Des in Indicates the conditions for a flow entity to enter a service node.

[0064] Des out Indicates the conditions under which a flow entity can leave a service node and proceed to the next service node.

[0065] The service node expresses the specific location that the material can reach in production, and is a mapping of its static spatial characteristics. The production execution activities in DMS are regarded as the dynamic flow of materials between service nodes, which reflects the dynamic spatial characteristics of materials. The execution logic is more complex, and an element is needed to describe the production activities at and between service nodes. Therefore, the concept of DSNP is proposed, which is defined as follows:

[0066] DSNP= <P p ,N 1 ,N 2 ,Dir,T l ,S e ,S p ,E> (2)

[0067] Among them, P p It is the type of DSNP, describing different production execution activities according to the type of service node and its connection relationship, such as Figure 3 shown.

[0068] N1 and N 2 There are two service nodes defined, representing the two locations in the physical space where production activities will take place.

[0069] Dir is N 1 and N 2 The direction of the connecting edge between them indicates the flow relationship of materials between service nodes. It can be bidirectional (0), positive (1, indicating N 1 To N 2 ) or reverse (2, N 2 Indicates to N 1 ).

[0070] T l It is the logistics time, which represents the time required to perform material transfer activities between two service nodes.

[0071] S e It is the status of the DSNP connection edge, including idle, working and blocked, which is affected by FICM and can reflect FICM.

[0072] S p It is the status of DSNP, including activated and inactivated, reflecting the execution process of production activities.

[0073] E is the executor of DSNP, which represents the logistics equipment that performs material transfer activities between two service nodes.

[0074] Two types of encapsulated service units ESC are defined, namely process encapsulation service unit PESC and buffer encapsulation service unit BESC, which are defined as follows:

[0075]

[0076] Among them, C, P and B are control devices, processing devices and cache devices respectively.

[0077] It is a collection of ESC actuators (logistics equipment) used to realize the material flow therein.

[0078] DSNP set It is a DSNP collection in ESC, which is used to describe the production execution logic of the unit.

[0079] LN set It is a collection of internal logistics paths of the ESC, which is used to realize the self-organization of actuators on the ESC path. For example, if an automatic warehouse is combined with a stacker as an ESC, the stacker track is the internal logistics path of the unit management.

[0080] and It is the infeed and outfeed location of the unit, representing the material interaction interface between the ESC and its external elements.

[0081] For the process encapsulation service unit PESC, and are its private input and output buffers.

[0082] For Buffered Encapsulated Service Units BESC, PESC set It is a collection of PESCs for which storage services can be provided. It is a common buffer between PESCs.

[0083] The production execution logic model PELM-DaE based on DSNP and ESC is defined as follows:

[0084]

[0085] Among them, PESC set and BESC set It is a collection of PESC and BESC, describing the composition of various production function modules in the workshop.

[0086] It is a workshop-level DSNP collection. and It is a collection of logistics paths and actuators at the workshop level, which together describe the interaction between ESCs in the workshop and are a mapping of the production activities in the workshop.

[0087] F set It is a flow entity set. It carries information such as process and production data, and drives the production organization and dynamic evolution of the workshop PELM-DaE.

[0088] Step 3: Construct process connectivity graph and operation connectivity graph. Based on PELM-DaE, the present invention further defines process and operation connectivity graphs to effectively describe the execution relationships and constraints between and within operations, and provide valuable input for scheduling and simulation.

[0089] The process connectivity graph is a set of DSNP sequences, representing the set of possible production activity execution routes for a product driven by its process, i.e., possible future FICM, which is defined as follows:

[0090] PCM={P 1 ,P 2 ,P i ,…,P n} (6)

[0091]

[0092] Among them, P i Indicates the DSNP sequence set contained in each process of the product, Indicates the current DSNP set, Indicates the set of subsequent DSNPs to which the current DSNP is connected.

[0093] Algorithm 1 gives the algorithm for constructing the process connectivity graph. Its input is the optional ESC information of each process in the product process and the constructed PELM-DaE. The algorithm continuously searches for a DSNP path that can be connected in the model and eventually reaches the ESC that terminates the process.

[0094] Algorithm 1 Process connectivity generation pseudo code

[0095]

[0096]

[0097] The operation connectivity graph is a subgraph of the process connectivity graph, which reduces the connectivity between a certain class of optional ESCs to a specific ESC; specifically, a process connectivity graph belongs to a certain class of products, and each material entity has an operation connectivity graph; its definition and construction process is similar to the process connectivity graph. Figure 1 To.

[0098] These two types of connectivity graphs have different application stages in the actual production process. For example, in the preparation stage before scheduling, the process connectivity graph can be used to express the connectivity between ESCs; in the simulation stage before production and during production, the job connectivity graph is used to map the scheduling results and simulate the production execution route; in the real-time monitoring stage during production, since it mainly involves the synchronous perception of FICM, both types of connectivity graphs can be used.

[0099] Step 4: Implement job scheduling based on the process connectivity diagram. Before job scheduling, it is necessary to build a process connectivity diagram, which can evaluate the rationality of the process route and the feasibility of production execution, thereby avoiding misjudgment of production capacity. If the processes of a certain product type cannot be connected in sequence, it indicates that the process design is defective and needs to be redesigned and evaluated.

[0100] Subsequently, scheduling calculations can be performed for materials evaluated through process connectivity to generate job plans. The process connectivity diagram expresses the execution relationships and constraints between and within jobs by pre-calculating FICM, which can be used as input for job scheduling and provide valuable insights.

[0101] For example, candidate ESCs can be recommended by evaluating the constraints between job operations in the process connection graph, such as logistics execution and resource utilization. The proposed framework supports the integration of existing scheduling algorithms. For each process of each material, candidate ESCs will be recommended based on the current constraints, and the scheduling algorithm will select one ESC from them for processing. After the scheduling is completed, the results will be mapped to the job connection graph of each material entity, thereby driving simulation verification and production execution.

[0102] Step 5: Implement production simulation based on the job connectivity graph. In order to test the feasibility of the scheduling results and determine whether the orders can be delivered on time according to the production plan, the production execution process needs to be simulated. It will be driven by a discrete event simulation algorithm, which advances each DSNP and simulation clock by inputting the job connectivity graph of each material and simulating production activity events until all orders are completed. The simulation algorithm is shown in Algorithm 2.

[0103] Algorithm 2 Production execution process simulation algorithm pseudo code

[0104]

[0105]

[0106] After the simulation is completed, if the predicted order cannot be completed on time, it is necessary to return to the scheduling calculation step to readjust the production plan. The verified scheduling results and the implementation path of the job connectivity graph will be input into the actual workshop as FICM's prediction to drive production execution.

[0107] Example:

[0108] 1. Actual production line analysis and production execution logic modeling.

[0109] The present invention has developed a corresponding software platform to help other researchers to implement the method of the present invention. Subsequently, a case study was conducted in a real workshop.

[0110] The workshop is called the Intelligent Manufacturing Demonstration Line (IMDL), which specializes in processing small and medium-sized structural parts. Figure 4 As shown. It is equipped with ten machine tools, two six-degree-of-freedom manipulators, three AGVs (one of which is equipped with a manipulator) and two automatic warehouses. It consists of five main areas: raw material and finished product storage area, coding and engraving area, two milling areas and finished product visual inspection area. All materials are coded by the laser engraving machine before processing, and quality inspection is carried out after processing before they can enter the finished product warehouse.

[0111] In order to implement the present invention, it is necessary to first model the production execution logic of the workshop. Figure 5 Key screenshots showing the IMDL graphical representation of PELM-DaE and the corresponding modeling performed in the platform.

[0112] It can be seen that IMDL contains 12 PESC: PESC set =[ESC i |1≤i≤12}, 2 BESC: BESC set ={ESC i |13≤i≤14}, 6 external logistics routes: 2 internal logistics routes: 7 external actuators: 2 internal actuators: 294 service nodes and 242 DSNPs.

[0113] 2. Build a process connectivity diagram and schedule jobs based on it.

[0114] IMDL currently produces 7 different types of products, and the process routes are shown in Table 1. Before scheduling the job, it is necessary to first evaluate the connectivity of the product process routes in the current order to determine whether the workshop can produce the various products in the order. Figure 6 As shown in the figure, by inputting the process route information of the current product into the constructed PELM-DaE, the platform automatically generates a process connectivity graph and assigns advancement times to service nodes and DSNP edges. Production managers can preview and simulate possible execution routes for products, which helps them evaluate the legitimacy of the current process route and make timely adjustments when equipment failure or other problems cause the process route to be disconnected.

[0115] Table 1 IMDL products and process routes

[0116]

[0117] For example, the process connectivity diagram for product 5 is as follows Figure 6 (b) As shown. It can be seen that due to ESC 3 The machine in is currently under maintenance, so DSNP 3-8 The process connectivity assessment of product 5 fails because it is in a disconnected state. Managers can use this information to delay the start of orders related to product 5 or design a new process route for it.

[0118] Next, we need to schedule the orders that have passed the process connectivity evaluation. Since the constructed process connectivity diagram realizes the pre-calculation of FICM, FICM expresses the execution relationship and constraints between and within processes, which can assist the job scheduling calculation process, eliminate unreasonable scheduling options, such as ESC disconnection caused by logistics partitioning or equipment failure, and recommend candidate units. The combination of the two aims to reduce the number of candidate ESCs in job scheduling, guide and simplify the scheduling calculation process, and improve the solution efficiency.

[0119] For example, Figure 6 (c) represents the process connectivity graph of product 7, assuming that its second process is scheduled as ESC 7 If the proposed model is not used, the scheduling algorithm needs to select one of the nine ESCs to perform the third process. However, in order to avoid interaction between AGVs, the logistics of production 1 and production 2 areas are currently disconnected, resulting in the inability to schedule subsequent processes to ESCs. 1 -ESC 5 The current scheduling algorithm cannot identify and avoid this situation due to the lack of consideration of fine-grained logistics. With the help of the information of the process connectivity graph, the ESC 1 -ESC 5 Eliminate and avoid unreasonable scheduling. At the same time, ESC can be further recommended based on the distance constraints between service points in the process connectivity graph. 6 and ESC 8 The above operation reduces the number of candidate ESCs to two, effectively reducing the difficulty and time consumption of the algorithm solution.

[0120] In order to further compare the improvement of the scheduling calculation efficiency by the proposed model, this embodiment uses a genetic algorithm to optimize the job scheduling of the workshop. Table 2 lists the average number of candidate ESCs and the average solution time of 100 scheduling runs before and after the application of the proposed model. The results show that the use of process connectivity diagrams to assist scheduling reduces the average number of candidate units by 54.4% and the average scheduling solution time by 23.9%. It can be expected that this advantage will be more obvious when the number of product operations increases. At the same time, the search space of the scheduling algorithm can be effectively reduced with the help of process connection diagrams. This enables the system to use more direct and efficient scheduling algorithms, and in some cases directly use manual decisions to achieve human-computer collaboration.

[0121] Table 2 Comparison of scheduling algorithm efficiency before and after using process connectivity graph

[0122]

[0123] 3. Build a job connectivity graph and perform production simulation based on it.

[0124] Since scheduling algorithms usually lack sufficient consideration of dynamic resource constraints such as logistics and caching in production, simulation is usually required to derive and verify scheduling results before production execution. Simulating the production execution process can help predict potential deviations between plans and actual execution, and evaluate the effects of different scheduling schemes and control logic, thereby helping managers select the optimal solution. In addition, during the production execution process, managers can track the production status through real-time simulation to ensure that the production process proceeds as planned. When abnormal events such as order delays are predicted, simulation can provide timely feedback to help managers make quick decisions and trigger rescheduling when necessary to ensure the continuity and efficiency of production.

[0125] Figure 7 The system interface for the production process simulation is shown. The current production order includes two products 4 and two products 6. Based on the PELM-DaE and job scheduling results, the developed platform generates a job connectivity graph for each material, representing the spatial information of the material in production. At the same time, according to the order of scheduling jobs, a material handling queue is generated for each ESC, representing the time information in production. This fusion of spatial and temporal data can provide a more comprehensive description of FICM in workshop production. Subsequently, this information is input into the simulation calculation service, and the process of production execution and FICM cycle is derived according to the preset control decision conditions and the simulation algorithm described in Algorithm 2. By recording the time when the state of elements in DSNP changes during the simulation calculation process, their process information such as running, idle and waiting time can be accumulated, thereby realizing the evaluation of simulation results and verification of scheduling schemes.

[0126] The input of traditional scheduling-type job-driven simulation often only includes the sequence and start and stop times of the jobs on the machine tools, while dynamically calculating the execution process of production activities such as logistics routes and in and out cache locations. By constructing a job connectivity graph as a simulation input, all possible production execution routes have been pre-calculated. In each simulation decision, the algorithm only needs to choose from possible material flow options based on current information and control flow to simulate production activities, without having to look for various possibilities before making decisions as before. This means improved simulation efficiency, which is crucial for timely production evaluation and optimization.

[0127] In order to verify this advantage, this embodiment compares the differences in running time of the production simulation algorithms based on the Seven-Element (SE) model, the Seven-Element (MNOSE) model for material nodes, and the connectivity graph model proposed in the present invention. Table 3 shows the average time for 100 simulation runs based on the production plan, and the scheduling plan is updated once every 10 simulation runs on average. The average number of workpieces for each simulation is 17. It should be pointed out that since the job connectivity graph is a subgraph of the process connectivity graph, its construction process is faster.

[0128] Table 3 Comparison of computational efficiency of three simulation methods

[0129]

[0130] The results show that the simulation computation time based on the proposed model is reduced by an average of 74.3% compared with the two models, while the total simulation time of 100 runs is reduced by an average of 67.18%.

Claims

1. An efficient scheduling and simulation method for discrete manufacturing workshops based on connectivity graph, characterized in that: The following steps are involved: Step 1: Define the workshop production execution logic model PELM-DaE based on directed service node pairs and encapsulated service units, formalize the execution logic of the discrete manufacturing workshop, and realize the integrated description of the workshop information flow, control flow and material flow FICM; Step 2: Construct a connectivity graph. Based on PELM-DaE, the process connectivity graph and the job connectivity graph are constructed respectively under the drive of the process route and the scheduling scheme to realize the flexible connectivity graph of the production execution logic, express the dynamic execution relationship and constraints between jobs, and then realize the pre-calculation of FICM to support efficient scheduling and simulation; Step 3: Scheduling and simulation based on the connectivity graph, by applying the process connectivity graph to job scheduling, the candidate equipment for job scheduling can be reduced; By applying the job connectivity graph to production simulation, the simulation computing efficiency and algorithm maintainability are improved.

2. The method for efficient scheduling and simulation of discrete manufacturing workshops based on connectivity graph according to claim 1 is characterized in that: The step 1 specifically includes: defining a directed service node pair DSNP and an encapsulated service unit ESC, and combining them to form a PELM-DaE, thereby realizing an effective expression of the workshop production execution logic; First, define the service node as follows: SN=<P n ,A n ,Pos,T s ,S n ,F set ,Des in ,Des out > (1) Among them, P n A is the service node type, which is divided into path node, processing node, storage node and auxiliary node according to the different services performed by the service node; A n is the attribute element of the service node, including logistics path, processing node, and buffer zone; Pos is the coordinate of the service node, indicating the specific coordinate position where the material arrives and stops during the production process; T s S is the service time, which means the time that the material stays in the service node and performs production activities after entering the service node; n is the status of the service node, including idle, working and blocked; F set It is the set of flow entities currently executing services in the node; in Indicates the conditions for the flow entity to enter the service node; Des out Indicates the conditions under which the flow entity can leave the service node and go to the next service node; The definition of directed service node pair DSNP is as follows: DSNP=<P p ,N1,N2,Dir,T l ,S e ,S p ,E> (2) Among them, P p is the type of DSNP, describing different production execution activities according to the type of service nodes and their connection relationships; N1 and N2 are two defined service nodes, representing the two locations in the material space where production activities will occur; Dir is the direction of the connection edge between N1 and N2, indicating the flow relationship of materials between service nodes, which can be bidirectional, forward or reverse; T l is the logistics time, which means the time required to perform material transfer activities between two service nodes; S e is the state of the DSNP connection edge, including idle, working and blocked; S p is the state of DSNP, including activated and inactivated, reflecting the execution process of production activities; E is the executor of DSNP, representing the logistics equipment that performs material transfer activities between two service nodes; Two types of encapsulated service units ESC are defined, namely process encapsulation service unit PESC and buffer encapsulation service unit BESC, which are defined as follows: Among them, C, P and B are control device, processing device and cache device respectively; It is a set of actuators of ESC, used to realize the material flow; DSNP set It is a DSNP set in ESC, which is used to describe the production execution logic of the unit; LN set It is the set of internal logistics paths of ESC, which is used to realize the self-organization of actuators on the ESC path; and It is the feed and discharge location of the unit, representing the material interaction interface between the ESC and its external elements; For the process encapsulation service unit PESC, and Is its private input and output buffer; for the buffer encapsulation service unit BESC, PESC set It is a collection of PESCs for which storage services can be provided; The production execution logic model PELM-DaE based on DSNP and ESC is defined as follows: Among them, PESC set and BESC set It is a collection of PESC and BESC, describing the composition of various production function modules in the workshop; It is a collection of DSNPs at the workshop level; and It is a collection of logistics paths and actuators at the workshop level, which jointly describe the interaction between ESCs in the workshop and are a mapping of the production activities in the workshop; F set It is a fluid entity set.

3. The method for efficient scheduling and simulation of discrete manufacturing workshops based on connectivity graph according to claim 2 is characterized in that: The step 2 of constructing the process connectivity diagram and the operation connectivity diagram is specifically as follows: The process connectivity graph is a set of DSNP sequences, representing the set of possible production activity execution routes for a product driven by its process, i.e., possible future FICM, which is defined as follows: PCM={P1,P2,P i ,…,P n } (6) Among them, P i Indicates the DSNP sequence set contained in each process of the product, Indicates the current DSNP set, Indicates the set of subsequent DSNPs to which the current DSNP is connected; The operation connectivity diagram is a subgraph of the process connectivity diagram, which reduces the connectivity between a certain type of optional ESCs to a specific ESC; specifically, a process connectivity diagram belongs to a certain type of product, and each material entity has an operation connectivity diagram; its definition and construction process are consistent with the process connectivity diagram.

4. The method for efficient scheduling and simulation of discrete manufacturing workshops based on connectivity graph according to claim 1 is characterized in that: In step 3, the operation scheduling is realized based on the process connectivity diagram as follows: Before scheduling operations, build a process connectivity diagram to evaluate the rationality of the process route and the feasibility of production execution, thereby avoiding misjudgment of production capacity; Subsequently, scheduling calculations are performed on the materials that have passed the process connectivity assessment to generate an operation plan; The process connection diagram expresses the execution relationships and constraints between and within jobs by pre-calculating FICM, which is used as input for job scheduling and provides valuable insights.

5. The method for efficient scheduling and simulation of discrete manufacturing workshops based on connectivity graph according to claim 1 is characterized in that: The specific implementation of production simulation based on the job connectivity graph in step 3 is as follows: In order to test the feasibility of the scheduling results and determine whether the orders can be delivered on time according to the production plan, the production execution process is simulated; it will be driven by a discrete event simulation algorithm, which advances each DSNP and simulation clock by inputting the job connection diagram of each material and simulating production activity events until all orders are completed; After the simulation is completed, if the predicted order cannot be completed on time, it is necessary to return to the scheduling calculation step to readjust the production plan; The verified scheduling results and the implementation path of the job connectivity graph will be input into the actual workshop as FICM predictions to drive production execution.

Images