Digital workshop planning and scheduling system risk assessment method
Through genetic algorithm and Q-learning algorithm, the workshop production scheduling scheme is optimized, and risk assessment is carried out in combination with hierarchical analysis method, which solves the problems of low production efficiency and high planning risks in digital workshops, achieving more efficient production and lower risk assessment.
Patent Information
- Application Number
- CN202411068064.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-22
AI Technical Summary
Due to the increasing complexity of equipment in the digital workshop, the original production planning arrangement is no longer the optimal solution, resulting in low production efficiency and the inability to obtain dynamic information of the processed products in real time, resulting in low production efficiency in the workshop. The existing planning plan needs to be systematically carried out. Assessment to reduce risk.
Using the workshop production scheduling optimization method based on genetic algorithm and Q-learning algorithm, the optimized workshop plan and scheduling plan are constructed, and the optimized workshop scheduling plan is obtained through risk assessment and analysis, and the operation plan completion rate, equipment utilization rate and maximum Processing completion time, overall risk assessment by hierarchical analysis method is carried out.
Improve production efficiency and reduce planning risks. Through the optimized workshop scheduling plan, shorter completion time and higher equipment utilization are achieved, reducing production risks and ensuring normal production of digital workshops.
Smart Images

Figure CN119005687A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a risk assessment method for a planning and scheduling system of a digital workshop, and belongs to the technical field of flexible manufacturing. Background Art
[0002] Workshop planning and scheduling is a crucial technology in the manufacturing system, which aims to reasonably allocate and plan resources within the workshop, ensure that production tasks are completed on time, maximize the use of resources, optimize the production process, and achieve efficient collaboration. With the continuous development of manufacturing technology, workshop planning and scheduling have become more and more critical, and need to cope with increasingly complex and changing production needs, meet diversified product portfolios and personalized production requirements. Therefore, workshop planning and scheduling has gradually evolved into flexible workshop planning and scheduling, introducing intelligent optimization algorithms, flexible manufacturing technologies, and innovative technologies such as cloud computing and distributed scheduling to adapt to changing situations. Flexible workshop planning and scheduling provides a strong impetus for the transformation and upgrading of the manufacturing industry through continuous innovation and application of new technologies, helping enterprises to achieve a more efficient, intelligent and environmentally friendly production model.
[0003] Risk assessment is also known as safety assessment. Its main purpose is to ensure the normal production and equipment safety of digital production lines. With the development of intelligent manufacturing technology, digital production lines are also developing towards high automation and high integration. The complexity of digital production lines has increased, and more and more large-scale mechanized equipment has been introduced into production line operations, and these equipment have a production sequence relationship. With the increase in the complexity of the equipment in the workshop, the original production plan of the original workshop may not be the optimal solution with the shortest processing time. Therefore, an optimization solution for the flexible workshop scheduling problem is needed. In the workshop scheduling stage, due to the different types and quantities of processed products, the workshop processing schemes are different. During the processing process, the dynamic information of the processed products cannot be obtained in real time, and the production progress cannot be observed during this process, which will lead to low workshop production efficiency. In order to improve production efficiency and reduce planning risks, it is necessary to systematically evaluate the existing plans and analyze the current plans. Enterprises can decide to optimize the plans based on the evaluation results. Summary of the invention
[0004] Purpose of the invention: In view of the problems and shortcomings in the prior art, the present invention provides a risk assessment method for the planning and scheduling system of a digital workshop to complete the risk assessment analysis of the production line and ensure the normal production of the digital production line in the workshop.
[0005] Technical solution: A risk assessment method for the planning and scheduling system of a digital workshop, based on an optimized workshop production scheduling plan, conducts a risk assessment on the planning and scheduling system of a digital workshop; it includes the following steps:
[0006] Step 1, obtaining workshop scheduling plan information;
[0007] Step 2: construct a genetic algorithm model and output the optimized workshop plan and scheduling scheme through the genetic algorithm model;
[0008] Step 3: Conduct risk assessment based on the optimized workshop plan and scheduling scheme; and obtain the risk assessment result.
[0009] In step 1, the workshop scheduling information is obtained: first, the workshop scheduling information of the enterprise production department is obtained according to the digital workshop, or the production order of the enterprise resource planning (ERP) system is automatically received through the interface, and the name of the product to be processed, the number of production, and the distribution of materials are obtained. According to the original production plan, the number of production equipment and the number of products that can be produced by each equipment at the same time, as well as the production plan output, are obtained.
[0010] In step 2, a genetic algorithm model is constructed: by obtaining the provided workshop production scheduling plan information, the workshop production scheduling plan method optimized by the genetic algorithm is used to output the optimized workshop scheduling plan for risk assessment. The following steps are included:
[0011] Step 2.1: Based on the acquired workshop scheduling information, complete the population initialization of the workshop scheduling genetic algorithm; the processing procedures of the workpiece on the machine and the processing machine serial number of each process of the workpiece are sorted into a chromosome (i.e., a set of workshop scheduling plans), and the first half and the second half of the chromosome are the processing procedures of the workpiece on the machine and the processing machine serial number of each process of processing the workpiece, respectively.
[0012] Furthermore, the Q-learning algorithm is used to initialize the population, with the goal of obtaining a better initial chromosome population and a better workshop production scheduling plan. The currently processed workpiece is taken as the current environmental state, the behavior of selecting the next workpiece is taken as the action taken, and the selected workpiece is the next new environmental state. Since the idle time generated by the two workpieces processed in sequence is a negative effect, the maximum idle time minus the idle time of the corresponding process is used as its reward value.
[0013] Furthermore, the R table and Q table are calculated:
[0014] The R value is the reward value of the environment based on the action feedback, and the R table is the R value table. The present invention uses the idle time of the workpiece for calculation, and the formula is as follows:
[0015]
[0016] R w,y =max(R)-I w,y
[0017] Among them, I w , u represents the idle time from workpiece w to workpiece y, C w , u It represents the maximum processing completion time from workpiece w to workpiece y, t y,k represents the processing time of workpiece y on machine k, R w,y Represents the reward value from artifact w to artifact y.
[0018] The Q value is the discounted value of the maximum cumulative reward obtained after selecting action a from state s and executing it. The Q table is the Q value table.
[0019] Q(s,a) new =(1-α)Q(s,a) old +α[R(s,α)+ymax[Q(s new , a new )]]
[0020] Q(s,a) old = = R (s, a) + γmaxQ (s new , a new )
[0021] Among them, α∈(0,1) is the learning rate, γ∈(0,1) is the discount rate, and s new It is the new state that state s reaches after selecting action a. Q(s, a) new is the Q value calculated from state s after selecting action a, Q(s,a)o ld It is the Q value calculated from state s before selecting action a, and R(s, a) is the reward value of state s when action a is selected.
[0022] The calculated Q table is used to guide the initialization process of the population and obtain the initial population. The remaining optional workpieces are placed in a temporary list for storing chromosomes. According to the value of the Q table, the workpiece with the largest corresponding Q value is selected as the next workpiece to be processed and removed from the temporary list. This step is repeated until the temporary list is empty to obtain a better individual. In this way, the better individuals of the corresponding number of workpieces can be obtained to form the initial population. In this process, the sequence of the processing procedures of the workpiece on the machine and the serial number of the processing machine for each process of the workpiece are respectively initialized according to the Q-learning algorithm. The better individuals obtained after the population is initialized form different m chromosomes, and the m chromosomes form a chromosome population M (that is, a workshop scheduling plan population M), m∈M.
[0023] Step 2.2: The objective function of the generated workshop scheduling plan is determined as the minimum completion time of all workpieces. The shorter the completion time of all workpieces, the better the workshop production scheduling plan is. The generated workshop scheduling plan meets specific conditions and its objective function formula is determined.
[0024] The objective function formula is:
[0025]
[0026] where f m Expressed as the objective function value of chromosome m, C i , j It is expressed as the time when processing workpiece i is completed on processing machine j of the last processing step, where J represents the set of all processing machines; the specific condition refers to the constraints imposed on these processes by the workpiece taking into account other processes, including processing sequence constraints on the same machine or process sequence constraints on the same workpiece.
[0027] Step 2.3: Randomly select two chromosomes from the population of all chromosomes, and take out the first half of each chromosome and the random part of the second half to cross. The initial iterative crossover probability of the pairwise crossover of chromosomes is p cmax The two parts are respectively called the processing sequence of the workpiece on the machine or the processing machine serial number of each process of the workpiece. After cross-pairing, new daughter chromosomes are generated, and the new daughter chromosomes form a new population.
[0028] The selection operation adopts the strategy of retaining excellent individuals in the population by sorting the objective function value. The objective function value of each chromosome individual is sorted from large to small, and the chromosomes with the first k objective function values are selected to replace the chromosomes with the last k objective function values. The crossover operation adopts the multi-point crossover method, and the probability of crossover is:
[0029]
[0030] where p cx is the probability of crossover of the current evolutionary generation, g is the current evolutionary generation, gmax is the maximum number of iterations, p cmax is the initial crossover probability.
[0031] Step 2.4: Select offspring chromosomes from the population after crossover, and then randomly select the mutation positions of the first half and the second half of the offspring chromosomes. The mutation probability of the offspring chromosomes is p v The processing sequence of the workpieces in the first half of the offspring chromosome and the processing machine numbers in the second half are swapped for the mutation positions. The diversity of the population is increased through mutation.
[0032] Step 2.5: Loop through steps 2.2 to 2.4 until the maximum number of iterations gmax is reached, and the optimal solution of the genetic algorithm is obtained. The optimal solution is output to obtain the optimized workshop scheduling plan.
[0033] Furthermore, the risk assessment of workshop planning and scheduling is completed based on the workshop scheduling plan optimized by genetic algorithm.
[0034] In the genetic algorithm model, crossover and mutation are the key to ensuring the quality of the solution. Crossover and mutation are achieved by changing the processing order of certain processes on the critical path block of the current solution (i.e., the current scheduling plan), thereby generating a series of neighborhood solutions, i.e., offspring scheduling plans. After evaluating these neighborhood solutions, the best offspring scheduling plan is selected as the current solution in the next iteration, and the above process is repeated until the number of iterations is reached and the genetic algorithm ends.
[0035] Output optimized workshop plan and scheduling plan: Construct a genetic algorithm model based on the above workshop plan and scheduling information, and output the optimized workshop scheduling plan after genetic algorithm optimization for risk assessment.
[0036] Step 3, conduct risk assessment based on the optimized workshop plan and scheduling plan: According to the workshop scheduling plan optimized by the genetic algorithm model, calculate the job plan completion rate, equipment utilization rate and maximum processing completion time risk assessment value respectively.
[0037] By combing the risk assessment process and results of the workshop's planning and scheduling system, the analytic hierarchy process is used to conduct an overall risk assessment of the digital workshop. First, a hierarchical model is constructed based on workshop planning and scheduling. There are target layer, criterion layer and factor layer, where the target layer is the risk assessment of the planning and scheduling system of the digital workshop, the criterion layer is the workshop planning and scheduling, and the factor layer is defined according to the criterion layer as:
[0038] The corresponding risk factor assessment values T1, T2, and T3 of the scheduling plan’s job plan completion rate, equipment utilization rate, and maximum processing completion time.
[0039]
[0040]
[0041]
[0042] The original scheduling plan has the following components: job plan completion rate e1, equipment utilization rate k1, and maximum processing completion time g1; the scheduling plan after genetic algorithm optimization has the following components: job plan completion rate e2, equipment utilization rate k2, and maximum processing completion time g2. The job plan completion rate is expressed as the completion rate of the scheduling plan within the planned time.
[0043] Furthermore, the weight of each factor is determined. The weight is the relative importance of the evaluation result in the risk assessment of the planning and scheduling system of the entire digital workshop. The weight can be adjusted accordingly according to the actual application needs. The evaluation result of the workshop planning and scheduling system risk assessment is obtained based on the evaluation index. The calculation method of the workshop planning and scheduling system risk assessment index is:
[0044]
[0045] In the formula, p b It is the weight factor of the factor layer of workshop planning and scheduling.
[0046] p1 is the weight factor of the risk factor assessment value T1 of the job plan completion rate, p2 is the weight factor of the risk factor assessment value T2 of the equipment utilization rate, and p3 is the weight factor of the risk factor assessment value T3 of the maximum processing completion time.
[0047] Furthermore, the risk level of the evaluation indicator is defined as:
[0048] Major risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0-0.25, and its risk level is major risk.
[0049] High risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0.25-0.5, and its risk level is high risk.
[0050] Medium risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0.5-0.75, and its risk level is medium risk.
[0051] Low risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0.75-1, and its risk level is low risk.
[0052] Get the risk assessment results and propose improvement measures and suggestions: Analyze the assessment results and use the Gantt chart to visualize the optimization results of the optimized scheduling plan. Further inform the enterprise production department or ERP system of the assessment value obtained from the risk assessment results, and give optimized workshop scheduling plans and scheduling suggestions. Perform scheduling in real time according to on-site execution, re-optimize the production schedule for the plans to be executed according to the production progress, inform the enterprise production department or ERP system of the completion rate of the operation plan, the utilization rate of the equipment, and the maximum processing completion time, improve production efficiency and reduce planning risks, and complete the risk assessment of the planning and scheduling system of the digital workshop. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] Figure 1 is a flow chart of workshop planning and scheduling evaluation and suggestion in an embodiment of the present invention;
[0054] Figure 2 is a flow chart of a genetic algorithm according to an embodiment of the present invention;
[0055] Figure 3 A schematic diagram of an original workshop scheduling plan according to an embodiment of the present invention;
[0056] Figure 4 Schematic diagram of a workshop scheduling plan after genetic algorithm optimization according to an embodiment of the present invention. DETAILED DESCRIPTION
[0057] The present invention is further explained below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent forms of modifications to the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.
[0058] like Figure 1 As shown, a risk assessment method for a planning and scheduling system of a digital workshop is described. The technical solution of the present invention will be clearly and completely described below in combination with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0059] First, we obtain the product name, production quantity, and material distribution status from the digital workshop. We also obtain the number of production equipment and the number of products that can be produced by each equipment at the same time, as well as the planned production output, based on the original production plan.
[0060] The workshop production scheduling planning method optimized by genetic algorithm of the present invention is implemented, and the implementation steps are as follows: Figure 2 As shown, the steps include the following steps performed in sequence:
[0061] In the first step, according to the acquired workshop production plan information, the processing sequence of the workpiece on the machine and the processing machine number of each process of the workpiece are respectively composed of a chromosome (i.e., a scheduling plan), and the first half and the second half of the chromosome are respectively the processing sequence of the workpiece on the machine and the processing machine number of each process of the workpiece. In the embodiment of the present invention, the processing sequence of the workpiece on the machine and the processing machine number of each process of the workpiece are respectively initialized and arranged into 200 different chromosomes through the Q-learning algorithm, and the 200 chromosomes constitute a chromosome population M.
[0062] Calculate the Q value and R value.
[0063] The R value is the reward value of the environment based on the action feedback, and the R table is the R value table. The present invention uses the idle time of the workpiece for calculation, and the formula is as follows:
[0064]
[0065] R w,y =max(R)-I w,y
[0066] Q(s,a) new =(1-α)Q(s,a) old +α[R(s,a)+γmax[Q(s new , a new )]]
[0067] Q(s,a) old = = R (s, a) + γmaxQ (s new , a new )
[0068] An original workshop scheduling plan in this embodiment is:
[0069] Processing sequence O of machine 0 5,1 , O 4,1 , O 8,1 , O 8,3 , O 3,3 , the processing sequence of machine 1 is O 1,1 , O 8,2 , O 4,3 , O 6,3 , the processing sequence of machine 2 is O 2,1 , O 6,1 , O 7,1 , O 2,2 , O 8,4 , the processing sequence of machine 3 is O 7,3 , O 3,2 , O 1,4 , the processing sequence of machine 4 is O 6,2 , O 1,2 , O 2,4 , the processing sequence of machine 5 is O 4,2 , O 5,3 , O 1,3 , O 6,4 , the processing sequence of machine 6 is O 5,2 , O 3,1 , O 2,3 , O 7,4 , O 3,4 , the processing sequence of machine 7 is O 7,2 , O 5,4 , O 4,4 Specific Gantt chart as follows Figure 3 .
[0070] Among them, i,h Represents the hth process of processing the i-th workpiece.
[0071] In the second step, the objective function of the generated workshop scheduling plan is determined as the minimum completion time of all workpieces. The shorter the completion time of all workpieces, the better the workshop production scheduling plan. The generated workshop scheduling plan meets specific conditions and determines its objective function formula.
[0072] The objective function formula is:
[0073]
[0074] where f m Expressed as the objective function value of chromosome m, C i,j It is expressed as the time when the workpiece i is finished on the last processing machine j, where J represents the set of all processing machines; the specific condition refers to the constraints imposed by other processes on these processes, including the processing sequence constraints on the same machine or the process sequence constraints on the same workpiece. m Represented as the objective function value of chromosome m.
[0075] The third step is to cross-pair the chromosomes to generate new daughter chromosomes, which form a new population.
[0076] In the embodiment of the present invention, the initial crossover probability p of the chromosome is determined cmax 95%. Take two chromosomes randomly and select random positions for crossover. The probability of crossover is:
[0077]
[0078] The fourth step is to randomly select the mutation position of the offspring chromosome and perform mutation operation on the chromosome to increase the diversity of the population through mutation.
[0079] In the embodiment of the present invention, the probability p of the offspring chromosome mutation is determined. v The chromosomes in the group are mutated one by one to obtain a new chromosome group.
[0080] The fifth step is to loop through the second to fourth steps until the number of iterations reaches 100, and the optimal solution of this genetic algorithm is obtained. The optimal solution is output and the optimized workshop scheduling plan is obtained.
[0081] The number of iterations determined in the embodiment of the present invention is 100. After the second to fifth steps of the genetic algorithm are cyclically executed to reach the number of iterations, the workshop scheduling plan that meets the processing requirements is screened out as the optimal solution of this genetic algorithm, and the workshop scheduling plan of this optimal solution is output as the optimized workshop scheduling plan.
[0082] The optimized workshop scheduling plan of this embodiment of the present invention is:
[0083] Processing sequence O of machine 0 8,1 , O 5,1 , O 8,3 , O 3,3 , the processing sequence of machine 1 is O 4,1 , O 1,1 , O 8,2 , O 4,3 , O 6,3 , the processing sequence of machine 2 is O 7,1 , O 2,1 , O 2,2 , O 8,4 , the processing sequence of machine 3 is O 3,2 , O 7,3 , O 1,4 , the processing sequence of machine 4 is O 6,1 , O 1,2 , O 1,3 , O 2,4 , the processing sequence of machine 5 is O 4,2 , O 5,3 , O 6,4 , the processing sequence of machine 6 is O 1,3 , O 5,2 , O 4,4 , O 2,3 , O 7,4 , O 3,4 , the processing sequence of machine 7 is O 7,2 , O 6,2 , O 5,4 Specific Gantt chart as follows Figure 4 .
[0084] Furthermore, the risk assessment of workshop planning and scheduling is completed based on the workshop scheduling plan optimized by genetic algorithm.
[0085] According to the processing sequence of each workpiece process in the workshop scheduling plan, the operation plan completion rate, equipment utilization rate and maximum processing completion time risk assessment value are calculated respectively.
[0086] By combing the risk assessment process and results of the workshop's planning and scheduling system, the analytic hierarchy process is used to conduct an overall risk assessment of the digital workshop. First, the workshop planning and scheduling is constructed into a hierarchical model. There are target layer, criterion layer and factor layer. The target layer is the risk assessment of the planning and scheduling system of the digital workshop, the criterion layer is the planning and scheduling, and the factor layer is defined according to the criterion layer as:
[0087] The corresponding risk factor assessment values T1, T2, and T3 of the scheduling plan’s job plan completion rate, equipment utilization rate, and maximum processing completion time.
[0088]
[0089]
[0090]
[0091] The original scheduling plan has a 100% completion rate, a 61.5% equipment utilization rate, and a maximum processing completion time of 25 minutes. The scheduling plan after genetic algorithm optimization has a 100% completion rate, a 73.2% equipment utilization rate, and a maximum processing completion time of 21 minutes. The completion rate of the scheduling plan is expressed as the completion rate of the scheduling plan within the planned time. It can be calculated that T1, T2, and T3 in the embodiment of the present invention are 1, 0.86, and 0.84 respectively;
[0092] Furthermore, the weight of each factor is determined. The weight is the relative importance of the evaluation result in the risk assessment of the planning and scheduling system of the entire digital workshop. The weight can be adjusted accordingly according to the actual application needs. The evaluation result of the workshop planning and scheduling system risk assessment is obtained based on the evaluation index. The calculation method of the workshop planning and scheduling system risk assessment index is:
[0093]
[0094] In the formula, p b It is the weight factor of the factor layer of workshop planning and scheduling.
[0095] p1 is the weight factor of the risk factor evaluation value T1 of the operation plan completion rate, p2 is the weight factor of the risk factor evaluation value T2 of the equipment utilization rate, and p3 is the weight factor of the risk factor evaluation value T3 of the maximum processing completion time. In the embodiment of the present invention, p1, p2, and p3 are 0.4, 0.3, and 0.3 respectively;
[0096] Furthermore, the risk level of the evaluation indicator is defined as:
[0097] Major risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0-0.25, and its risk level is major risk.
[0098] High risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0.25-0.5, and its risk level is high risk.
[0099] Medium risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0.5-0.75, and its risk level is medium risk.
[0100] Low risk: The risk assessment index C of the workshop's planning and scheduling system is in the range of 0.75-1, and its risk level is low risk.
[0101] According to the calculation, the risk assessment index C of the workshop's planning and scheduling system is 0.868, which is a low risk.
[0102] Analyze the evaluation results and visualize the optimized scheduling plan using a Gantt chart. Further inform the enterprise production department or ERP system of the evaluation value obtained from the risk assessment results, and give optimized workshop scheduling plans and scheduling suggestions. Perform scheduling in real time according to on-site execution, re-optimize the production schedule for the plans to be executed according to the production progress, inform the enterprise production department or ERP system of the completion rate of the operation plan, the utilization rate of the equipment, and the maximum processing completion time, improve production efficiency and reduce planning risks, and complete the risk assessment of the planning and scheduling system of the digital workshop.
Claims
1. A risk assessment method for a planning and scheduling system of a digital workshop, characterized in that: Based on the optimized workshop production scheduling plan, the risk assessment of the planning and scheduling system of the digital workshop is carried out; the steps include: Step 1, obtaining workshop scheduling plan information; Step 2: construct a genetic algorithm model and output the optimized workshop plan and scheduling scheme through the genetic algorithm model; Step 3: Conduct risk assessment based on the optimized workshop plan and scheduling scheme; Obtain risk assessment results.
2. The risk assessment method for the planning and scheduling system of a digital workshop according to claim 1 is characterized in that: In step 1, the workshop scheduling plan information is obtained: first, the workshop scheduling plan information of the enterprise production department is obtained according to the digital workshop or the production order of the enterprise resource planning system is automatically received through the interface to obtain the name of the product to be processed, the production quantity, and the material distribution status. According to the original production plan, obtain the number of production equipment, the number that each equipment can produce at the same time, and the planned production output.
3. The risk assessment method for the planning and scheduling system of a digital workshop according to claim 1 is characterized in that: In the step 2, a genetic algorithm model is constructed: obtaining workshop production scheduling plan information, using a workshop production scheduling plan method optimized by a genetic algorithm, and outputting an optimized workshop scheduling plan for risk assessment; the steps include: Step 2.1: According to the acquired workshop scheduling plan information, the population of the workshop scheduling genetic algorithm is initialized; the processing procedures of the workpiece on the machine and the processing machine serial number of each process of the workpiece are sorted into a chromosome, and the first half and the second half of the chromosome are the processing procedures of the workpiece on the machine and the processing machine serial number of each process of the workpiece respectively; The population is initialized using the Q-learning algorithm. The currently processed workpiece is taken as the current environmental state, the behavior of selecting the next workpiece is taken as the action to be taken, the selected workpiece is the next new environmental state, and the maximum idle time minus the idle time of the corresponding process is taken as its reward value. Step 2.2: The objective function of the generated workshop scheduling plan is determined as the minimum completion time of all workpieces; and the generated workshop scheduling plan meets specific conditions, and its objective function formula is determined; Specific conditions refer to the constraints imposed on these processes by other processes on the workpiece, including processing sequence constraints on the same machine or process sequence constraints on the same workpiece. Step 2.3: Randomly select two chromosomes from the population of all chromosomes, and take out the first half of each chromosome and the random part of the second half to cross. The initial iterative crossover probability of the pairwise crossover of chromosomes is p cmax ; After cross-pairing, new offspring chromosomes are generated, and the new offspring chromosomes form a new population; The selection operation adopts the strategy of retaining excellent individuals in the population by sorting the objective function value. The objective function value of each chromosome individual is sorted from large to small, and the chromosomes with the first k objective function values are selected to replace the chromosomes with the last k objective function values. Step 2.4: Select offspring chromosomes from the population after crossover, and then randomly select the mutation positions of the first half and the second half of the offspring chromosomes. The mutation probability of the offspring chromosomes is p v and interchange the mutation positions of the processing sequence of the workpieces of the first half of the daughter chromosome and the processing machine numbers corresponding to the second half; Step 2.5: Loop through steps 2.2 to 2.4 until the maximum number of iterations is reached, and the optimal solution of the genetic algorithm is obtained. The optimal solution is output, and the optimized workshop scheduling plan is obtained.
4. The risk assessment method for the planning and scheduling system of a digital workshop according to claim 1 is characterized in that: The step 3 is to perform risk assessment based on the optimized workshop plan and scheduling scheme: based on the workshop scheduling plan optimized by the genetic algorithm model, calculate the operation plan completion rate, equipment utilization rate and maximum processing completion time risk assessment value respectively.
5. The risk assessment method for the planning and scheduling system of a digital workshop according to claim 1 is characterized in that: In step 3, the analytic hierarchy process is used to conduct an overall risk assessment of the digital workshop. First, a hierarchical model is constructed based on workshop planning and scheduling. The hierarchical model has a target layer, a criterion layer, and a factor layer, wherein the target layer is the risk assessment of the planning and scheduling system of the digital workshop, the criterion layer is the workshop planning and scheduling, and the factor layer is defined according to the criterion layer as follows: The corresponding risk factor assessment values T1, T2, and T3 of the operation plan completion rate, equipment utilization rate, and maximum processing completion time of the scheduling plan; The original scheduling plan has the following characteristics: the job plan completion rate e1, the equipment utilization rate k1, and the maximum processing completion time g1. The scheduling plan after genetic algorithm optimization has the following characteristics: the job plan completion rate e2, the equipment utilization rate k2, and the maximum processing completion time g2. The job plan completion rate is expressed as the completion rate of the scheduling plan within the planned time. Determine the weight of each factor. The weight is the relative importance of the evaluation result in the risk assessment of the planning and scheduling system of the entire digital workshop. The weight can be adjusted accordingly according to the actual application needs. The evaluation result of the workshop planning and scheduling system risk assessment is obtained based on the evaluation indicators. The calculation method of the workshop planning and scheduling system risk assessment indicators is: In the formula, p b It is the weight factor of the factor layer of workshop planning and scheduling. p1 is the weight factor of the risk factor assessment value T1 of the job plan completion rate, p2 is the weight factor of the risk factor assessment value T2 of the equipment utilization rate, and p3 is the weight factor of the risk factor assessment value T3 of the maximum processing completion time.
6. The risk assessment method for the planning and scheduling system of a digital workshop according to claim 1 is characterized in that: In step 2, the R table and the Q table are calculated: The R value is the reward value given by the environment based on the action feedback. The R table is the R value table, which is calculated using the idle time of the workpiece. The formula is as follows: R w,y =max(R)-I w,y Among them, I w,y represents the idle time from workpiece w to workpiece y, C w,y It represents the maximum processing completion time from workpiece w to workpiece y, t y,k represents the processing time of workpiece y on machine k, R w,y represents the reward value from artifact w to artifact y; The Q value is the discounted value of the maximum cumulative reward obtained after selecting action a from state s. The Q table is the Q value table; Q(s,a) new =(1-α)Q(s,a) old +α[R(s,a)+γmax[Q(s new ,a new )]] Q(s,a) old ==R(s,a)+γmaxQ(s new ,a new ) Among them, α∈(0,1) is the learning rate, γ∈(0,1) is the discount rate, and s new It is the new state that state s reaches after selecting action a; Q(s,a) new is the Q value calculated from state s after selecting action a, Q(s,a) old is the Q value calculated from state s before selecting action a, and R(s,a) is the reward value of state s when action a is selected; The calculated Q table is used to guide the population initialization process and obtain the initial population; the remaining optional workpieces are placed in a temporary list for storing chromosomes, and according to the value of the Q table, the workpiece with the largest corresponding Q value is selected as the next workpiece to be processed and removed from the temporary list until the temporary list is empty to obtain a better individual. In this way, the better individuals of the corresponding number of workpieces can be obtained to form the initial population; in this process, the processing sequence of the workpiece on the machine and the processing machine serial number of each process of the workpiece are respectively initialized according to the Q-learning algorithm. The better individuals obtained after the population is initialized form different m chromosomes, and the m chromosomes form a chromosome population M, m∈M.
7. The risk assessment method for the planning and scheduling system of a digital workshop according to claim 3 is characterized in that: In step 2.2, the objective function formula is: where f m Expressed as the objective function value of chromosome m, C i,j It is the time when workpiece i is finished on machine j in the last processing step, and J represents the set of all processing machines.
Citation Information
Cited By
Electronic equipment production plan compilation method and device
CN120317687A
Method and device for compiling production plan of electronic equipment
CN120317687B
