A Modeling and Scheduling Method for High-end Battery Production Model in Mass Customization

Abstract

The present invention discloses a method for modeling and scheduling the production of a large number of customized high-end batteries, belonging to the field of modeling and scheduling of power battery production models. The method analyzes the production processes of batteries, obtains an objective function aiming at maximizing battery profits, initializes the population, generates a sorting scheme by binary coding, sets a fitness function aiming at minimizing the completion time, performs selection, crossover, and mutation operations of the genetic algorithm, obtains the shortest completion time and sorting method, solves the contradiction between the mass production of products and the handling of customer customization requirements faced by battery enterprises, improves the production efficiency of batteries, and reduces the production cost.

Description

Technical Field

[0001] The present invention relates to a method for modeling and scheduling the production of high-end batteries in mass customization, and belongs to the field of modeling and scheduling of power battery production models. Background Art

[0002] As an efficient and lightweight battery energy source, the development speed of lithium batteries has far exceeded people's expectations. With advantages such as high energy density, no memory effect, and high tolerance, it has become the representative of the new generation of green energy batteries. With the development of lithium battery energy, electric vehicles have also been accepted and chosen by more and more users.

[0003] With the continuous maturity of the market, the production models of electric vehicle enterprises have gradually changed to customer-centered. Driven by this, the production mode of lithium battery production enterprises has gradually changed from large-scale mass production to large-scale customization production, so as to meet the personalized needs of electric vehicle enterprises for batteries.

[0004] In the large-scale mass production mode, since the produced batteries have the same specifications, that is, the parameters of each process are the same, the production efficiency can reach a relatively high level; while in the customization production mode, the time required for the same production process corresponding to different types of batteries is different. When a lithium battery production enterprise receives multiple customized orders, it is necessary to rationally design each process of different types of batteries, otherwise it will lead to low production efficiency of lithium battery personalization. Summary of the Invention

[0005] In order to improve the production efficiency of high-end batteries in mass customization, this application studies the construction and optimization of lithium battery production models. By taking into account the transfer time between each process in the battery production process (the transfer time corresponding to different processes is also different), and through improving the engineering depth indication algorithm combined with the genetic algorithm, the production efficiency of the battery is improved and the production cost is reduced.

[0006] A method for modeling and scheduling the production of high-end batteries in mass customization, the method includes:

[0007] Step S1: For the scheduling problem of batch production enterprises, improve the engineering depth indication algorithm, take into account the time required for equipment maintenance, and calculate the production lead time of lithium battery products based on the improved engineering depth indication algorithm;

[0008] Step S2: Consider the processing time and transfer time required for different processes, and establish a production model with the maximum profit of lithium batteries as the goal by analyzing the production process of batteries;

[0009] Step S3: Initialize the population using a genetic algorithm model, set the scheduling plan, and set the fitness function of the genetic algorithm to the production consumption time;

[0010] Step S4: Perform the crossover operation of the genetic algorithm according to the characteristics of high-end lithium battery production;

[0011] Step S5: Perform the mutation operation of the genetic algorithm according to the characteristics of high-end lithium battery production;

[0012] Step S6: Perform the selection operation of the genetic algorithm;

[0013] Step S7: Give the production scheduling plan and processing time according to the results of the genetic algorithm, and determine the final extraction period of the project and give the Gantt chart according to the production quantity and production situation.

[0014] Optionally, the step S1 includes:

[0015] Step S1.1, calculate the stock and maintenance time t ij , and its expression is:

[0016] t ij = k ij (a ij + b ij + c ij )z ij

[0017] where t ij is the total time required for the j-th process of the i-th lithium battery; a ij is the processing time of the i-th lithium battery in the j-th process; b ij is the corresponding transfer time, that is, the time for the i-th lithium battery to wait to enter the j-th process after completing the (j - 1)-th process; c ij is the time required for the equipment routine maintenance corresponding to the j-th process; k ij is the safety stock coefficient, and z ij is the prediction and corrective maintenance coefficient;

[0018] Step S1.2, calculate the insurance stock of the lithium battery, and its expression is:

[0019] s ij = t ij × Q i / h + W ij

[0020] where Q i is the daily production capacity of the i-th lithium battery; h is the daily working hours; W ij is the insurance stock data;

[0021] Step S1.3, calculate the total stock, and its expression is

[0022] s i = ∑s ij

[0023] Step S1.4, calculate the production lead time of lithium batteries:

[0024]

[0025] According to the inventory situation, obtain the production lead time T of the i-th type of lithium battery i .

[0026] Optionally, the production model established in step S2 with the goal of maximizing the profit of lithium batteries is:

[0027]

[0028] where, in a certain time period, N types of batteries are produced, λ i is the selling unit price of the i-th type of lithium battery; o i is the final production volume of lithium battery i; v ij represents the production cost of the i-th type of lithium battery in the j-th process; l ij is the storage cost required for the i-th type of lithium battery in the j-th process; q jm is the consumption of the m-th raw material in the j-th process; w jm is the cost of the m-th raw material in the j-th process; β i is the deferred payment cost; d i is the contractually stipulated delivery volume of the i-th type of lithium battery.

[0029] Optionally, the fitness function in step S3 is:

[0030] fitness(i,j) = eT ij

[0031] where, eT ij = sT ij + PT ij , eT ij represents the end time of the i-th type of lithium battery in the j-th process, sT ij is its start time, and PT ij is its total processing time.

[0032] Optionally, the crossover operation in the genetic algorithm adopts the two-point crossover method.

[0033] Optionally, the mutation operation in the genetic algorithm adopts the reverse order method.

[0034] Optionally, the selection operation in the genetic algorithm adopts the roulette wheel selection method.

[0035] Optionally, the method performs production scheduling for different types of battery orders.

[0036] This application also provides the application of the above method in battery production management.

[0037] The beneficial effects of the present invention are as follows:

[0038] The core objective of this application is to obtain the sorting method and the minimum completion time optimized by the genetic algorithm, that is, based on the improved engineering depth indication algorithm, combined with the genetic algorithm, continuously iterate the given fitness function to obtain the optimal solution, and optimize the genetic algorithm under the indication of the process inventory safety stock and the process production lead time, which greatly reduces the processing time and increases the profit. By using this production model and scheduling method, the production efficiency of the battery is improved, the production cost is reduced, and it provides guidance for modern battery manufacturing enterprises to realize mass customization production of products. Description of the Drawings

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0040] Figure 1 It is a basic process diagram of battery production disclosed in an embodiment of the present invention.

[0041] Figure 2 It is a flowchart of a method for modeling and scheduling the production of high-volume customized high-end batteries disclosed in an embodiment of the present invention.

[0042] Figure 3 It is a comparison simulation diagram between the algorithm of this application and the iterative algorithm.

[0043] Figure 4 It is a Gantt chart corresponding to an application example using the method of this application. Detailed Embodiments

[0044] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will further describe the embodiments of the present invention in detail in conjunction with the drawings.

[0045] As Figure 1 shown, the production process of the battery can generally be divided into three stages according to the workshop division: cell production, module assembly, and PACK final assembly. Each stage can be further divided into multiple processes, and there is a sequential order between the processes corresponding to the same type of battery.

[0046] For standard battery products in mass production, the parameters of each process are the same. Therefore, the produced standard battery products all have the same specification requirements. However, the personalized needs of customers may have different requirements for the batteries, that is, the parameters of the same process corresponding to different types of batteries are different. Taking the middle-stage module assembly stage as an example, according to the personalized needs of customers, after the production of battery cells is completed, multiple battery cells are selected according to the customized requirements of the order, and are assembled into a battery module through series and parallel combinations, and finally personalized PACK general assembly is carried out. The purpose of this application is to use an algorithm to optimize the sequencing between the production processes of different types of batteries according to different personalized needs of customers, so as to greatly improve production efficiency, reduce costs and increase profits.

[0047] As Figure 1 shown, the production process of the battery is divided into three stages: battery cell production, module assembly, and PACK general assembly; among them, the battery cell production stage involves sheet making process, assembly process, and formation process; the sheet making process includes the following processes: mixing process, coating process, and sheet making process; the assembly process includes the following processes: winding, stacking process, encapsulation, baking process, and liquid injection, sealing process; the formation process includes formation process, grading process, and testing process; the module assembly stage includes battery cell selection process, battery cell stacking process, and welding process; the PACK general assembly stage includes box assembly process, testing process, and final assembly process.

[0048] Embodiment 1:

[0049] This embodiment provides a method for modeling and scheduling the production of high-end batteries in mass customization, including:

[0050] Step S1: For the scheduling problem of batch production enterprises, improve the engineering depth indication algorithm, take into account the time required for equipment maintenance, and calculate the production lead time of lithium battery products based on the improved engineering depth indication algorithm.

[0051] Step S1.1, calculate the inventory and maintenance time t ij , and its expression is:

[0052] t ij = k ij (a ij + b ij + c ij )z ij

[0053] Among them, t ij is the total time required for the jth process of the ith lithium battery; a ij is the processing time of the ith lithium battery in the jth process; b ijis the corresponding transfer time, that is, the time when the i-th lithium battery waits to enter the j-th process after completing the (j - 1)-th process; c ij is the time required for the equipment routine maintenance corresponding to the j-th process.

[0054] It should be noted that if the parts processed in a process directly enter the next process after completing the previous process, then the transfer time b ij is zero at this time.

[0055] k ij is the safety stock coefficient, and its size depends on the process buffer time that the entire processing workshop can withstand. When a process does not need to wait and directly enters the next process, the safety stock coefficient is 1, which is the ideal situation for the process flow at this time.

[0056] z ij is the prediction and corrective maintenance coefficient, which refers to the time for maintenance when abnormal or emergency interruptions occur in the operation data of the workshop machines. Its size is determined by the ratio of the total consumption time of prediction and corrective maintenance to the total operation time in the past year, that is θ is the daily working hours of the workshop, X represents the average time of each maintenance in the workshop in the past year, and R represents the total number of maintenance times in the workshop in the past year. Step S1.2, calculate the insurance stock of lithium batteries, and its expression is:

[0057] s ij = t ij × Q i / h + W ij

[0058] Among them, Q i is the daily production capacity of the i-th lithium battery; h is the daily working hours; W ij is the insurance stock data. According to the stock time situation of the process and combined with the daily production capacity of the workshop, the process insurance stock data of the workshop can be calculated..

[0059] Step S1.3, calculate the total stock, and its expression is

[0060] s i = ∑s ij

[0061] By summing up all the process insurance stocks, the inventory capacity situation of all processes can be obtained.

[0062] Step S1.4, calculate the production lead date of lithium batteries

[0063]

[0064] According to the stock situation, the production lead date T of this lithium battery can be obtained i .

[0065] Step S2: Since the processing time and transfer time required for different processes are different, by analyzing the production process of the battery, a production model with the maximization of the profit of lithium batteries as the goal is established.

[0066]

[0067] Among them, within a certain time period, N types of batteries are produced, and λ i is the selling unit price of the i-th type of lithium battery; o i is the final production volume of lithium battery i; v ij represents the production cost of the i-th type of lithium battery in the j-th process; l ij is the storage cost required for the i-th type of lithium battery in the j-th process; q jm is the consumption of the m-th raw material in the j-th process; w jm is the cost of the m-th raw material in the j-th process; β i is the deferred payment cost; d i is the contractually stipulated delivery volume of the i-th type of lithium battery. All the above parameters are greater than zero.

[0068] Step S3: Adopt the genetic algorithm model, initialize the population, and set the scheduling plan;

[0069] In order to ensure the lowest production consumption time of the entire algorithm, the fitness function is set as the production consumption time.

[0070] The fitness function is the core of the entire algorithm and is also the condition for judging whether the individuals in the population are excellent, reflecting the idea of survival of the fittest in the algorithm. The fitness set in this application is the makespan, and the goal is to minimize the makespan. eT ij = sT ij + PT ij , where eT ij represents the end time of the i-th type of lithium battery in the j-th process, sT ij is its start time, and PT ij is its total processing time. Fitness function:

[0071] fitness(i,j) = eT ij

[0072] Input the algorithm control parameters, that is, the working hours and equipment parameters data, form the binary encoding chorme on the basis of Step S1, and form the scheduling plan according to the encoding chorme.

[0073] Define a two-dimensional array schdule with job * mach rows and 5 columns to store the scheduling scheme corresponding to the encoding. job * mach is determined by the process j and the number of devices K, enabling it to return the number of rows and columns of a specific array variable. Use the Size function to enable the array to return the number of rows and columns of the array variable. Fill the array schdule according to the arrays schdule and data. Define a one-dimensional array of length job to store the earliest start time of the current process of each job, or the completion time of the previous process of the current job. Define a one-dimensional array of length job to store the process number of the current scheduled production for each job. Then arrange the specific start and completion times corresponding to chorme in the schdule array.

[0074] Step S4: According to the characteristics of high-end lithium battery production, use two-point crossover to perform the crossover operation of the genetic algorithm.

[0075] The crossover operation and the mutation operation are operations that ensure the flexibility and global nature of the algorithm. The crossover operation is a recombination of genes. Through a certain crossover probability p c , part of the chromosomes of the population individuals can have partial gene crossover and exchange, thereby generating new individuals, improving the diversity of the population individuals, and also ensuring the global nature of the algorithm solution. In this application, the two-point crossover method is adopted. Two-point crossover means that two crossover points are randomly set in the individual chromosome, and then partial gene exchange is performed. The specific process of two-point crossover includes: randomly setting two crossover points in the coding strings of two paired individuals and exchanging the partial chromosomes between the two set crossover points of the two individuals.

[0076] Step S5: According to the characteristics of high-end lithium battery production, use the reverse order method to perform the mutation operation of the genetic algorithm.

[0077] The mutation operation is also a way to generate new individuals in the population. Through the setting of the mutation probability p m , the individuals in the population have a probability of replacing the parental genes with other genes, thereby forming new population individuals. The mutation operation increases the diversity of the species and improves the search ability of the algorithm. The setting of the mutation factor needs to be within a certain range. If the range is too small, no new individuals will be generated and it will fall into local optimality. If the range is too large, the search process for the optimal solution will lack heredity and become a random algorithm, affecting the performance. In this application, the reverse order method is adopted. The reverse order method comes from the determinant. In a permutation, if the front and back positions of a pair of numbers are opposite to the size order, that is, the number in front is greater than the number behind, then they are called an inversion.

[0078] Step S6: Perform the selection operation of the genetic algorithm.

[0079] The selection operation is a kind of replication in a sense. Selecting the superior individuals from the population and eliminating the inferior individuals is called selection. The selection operator is sometimes also called the reproduction operator. The purpose of selection is to directly inherit the optimized individuals to the next generation or generate new individuals through pairing and crossover and then inherit them to the next generation. In this right, the roulette wheel selection method is adopted. The roulette wheel method is the simplest and most commonly used selection method. Its basic idea is that the selection probability of each individual is proportional to its fitness value. The probability reflects the proportion of the fitness of an individual in the total fitness of all individuals in the population. The greater the individual fitness, the greater the probability of its being selected, and vice versa. After calculating the selection probabilities of each individual in the population, in order to select mating individuals, multiple rounds of selection are required. Each round generates a uniform random number between [0, 1], and this random number is used as a selection pointer to determine the alternative individuals.

[0080] When the fitness of the best individual reaches the given threshold, or the fitness of the best individual and the population fitness no longer increase, or the number of iterations reaches the preset number of generations, the algorithm terminates and outputs the optimal solution obtained by the algorithm.

[0081] Step S7: Give the production scheduling plan and processing time, and determine the final extraction period of the project and give a Gantt chart according to the production quantity and production situation.

[0082] Draw a Gantt chart according to the production scheduling plan, beautify the Gantt chart design, and output the best sorting plan and the maximum completion time.

[0083] The production process of the battery is divided into three stages: cell production, module assembly, and PACK final assembly; among them, the cell production stage involves sheet making process, assembly process, and formation process; the sheet making process includes the following processes: batching process, coating process, and slitting of positive and negative electrodes process; the assembly process includes the following processes: winding, stacking process, encapsulation, baking process, and injection, sealing process; the formation process includes formation process, grading process, and testing process;

[0084] The module assembly stage includes cell selection process, cell stacking process, and welding process;

[0085] The PACK final assembly stage includes box assembly process, testing process, and final assembly process.

[0086] To verify the effectiveness of the method of this application, taking the actual production requirements of a lithium battery cell and PACK system manufacturing enterprise as an example, it is described as follows:

[0087] A lithium battery cell and PACK system manufacturing enterprise schedules the production plans of cells and PACK systems. As a key component of the PACK system, the cell needs to be put into production in advance according to the lead time calculation. The PACK system is put into production in advance according to the customer order requirements, and the PACK system conducts mass customization assembly for orders according to the customer order requirements. Suppose there are 6 different models of products mixed and assembled on the production line, namely products of types A, B, C, D, E, and F. The production capacity of a certain type of cell in the production workshop is 10,000 per day, and the daily working hours are 20h. The other processing time and transfer time of the process are shown in Table 1 below.

[0088] For a manufacturing enterprise, profit is its main pursuit goal. Therefore, a mathematical model is constructed with profit maximization as the production model and profit maximization as the production goal. The expression of the objective function:

[0089]

[0090] To ensure the shortest overall production time of the algorithm, the fitness function is set to the consumption time, and the expression is: f(v k )=c max , where C max is the maximum flow time of the plan.

[0091] Table 1

[0092]

[0093] Set the population size to 100, the crossover probability to 0.4, and the mutation probability to 0.8. Set the maximum number of algorithm iterations to 300 times, and use simulation software to verify the algorithm.

[0094] To highlight that the sorting result of the algorithm in this application can save processing time, the improved algorithm in this application is compared with the actual experience method, and different algorithms are used to optimize the product processing and assembly process of lithium batteries. The results are shown in Table 2.

[0095] Table 2: Comparison of different sorting schemes

[0096]

[0097] It can be seen from the results in Table 2 that after optimization by different algorithms, the obtained processing sequences are different. Comparing the two methods, the actual experience processing method requires longer processing time, while the improved method in this application has shorter processing time, and the time-consuming is 245.7h.

[0098] Using different algorithms, obtain the profit situation of this production task, as Figure 3 shown, by Figure 3It can be seen that for the case of the profit of the entire production, the method of the present application is due to the actual experience method. The actual experience method adopts manual arrangement, so its profit maintains a horizontal straight line, while the method of the present application first increases with the number of iterations and then tends to be stable, obtaining a stable value. Figure 4 is a Gantt chart corresponding to an application example of the method of the present application, Figure 4 in the horizontal axis direction represents time, and the vertical axis direction represents the machine number. In [x y], x represents the process number job, and y represents the machine number. The corresponding line represents the start time and end time of this operation (since the times corresponding to each process vary greatly, Figure 4 there is a problem of overlapping annotations of some [x y] values in the figure, and the overlapping [x y] values are marked in the corresponding dashed box).

[0099] After determining the processes of the product, in order to ensure the smooth supply of the product, it is also necessary to ensure the preparation and supply of the product's supporting materials before production. According to the requirements of the improved engineering depth indication algorithm of the present application, the production of the battery cells of lithium batteries requires a certain lead time and inventory. As shown in Table 3

[0100] Table 3 Situation of process insurance inventory and advance production scheduling date

[0101]

[0102] Some steps in the embodiments of the present invention can be implemented by software, and the corresponding software program can be stored in a readable storage medium, such as an optical disc or a hard disk, etc.

[0103] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for modeling and scheduling in a high-volume customized high-end battery production model, characterized in that The method includes: Step S1: For the scheduling problem of mass production enterprises, improve the engineering depth indication algorithm, take into account the time required for equipment maintenance, and calculate the production lead time of lithium battery products based on the improved engineering depth indication algorithm; Step S2: Consider the processing time and transfer time required for different processes, and establish a production model with the maximization of lithium battery profit as the goal by analyzing the production process of batteries; Step S3: Adopt a genetic algorithm model, initialize the population, set the scheduling plan, and set the fitness function of the genetic algorithm as the production consumption time; Step S4: According to the characteristics of high-end lithium battery production, perform the crossover operation of the genetic algorithm; Step S5: According to the characteristics of high-end lithium battery production, perform the mutation operation of the genetic algorithm; Step S6: Perform the selection operation of the genetic algorithm; Step S7: Give the production scheduling plan and processing time according to the results of the genetic algorithm, and determine the final extraction period of the project and give a Gantt chart according to the production quantity and production situation; The said Step S1 includes: Step S1.1, calculate the stock quantity and the maintenance time t ij , and its expression is: t ij = k ij (a ij + b ij + c ij )z ij Among them, t ij is the total time required for the j-th process of the i-th lithium battery; a ij is the processing time of the i-th lithium battery in the j-th process; b ij is the corresponding transfer time, that is, the time for the i-th lithium battery to wait to enter the j-th process after completing the (j - 1)-th process; c ij is the time required for the equipment routine maintenance corresponding to the j-th process; k ij is the safety stock coefficient, z ij is the prediction and corrective maintenance coefficient; Step S1.2, calculate the safety stock of lithium batteries, and its expression is: s ij = t ij × Q i / h + W ij Among them, Q i is the daily production capacity of the i-th type of lithium battery; h is the daily working hours; W ij is the safety stock data; Step S1.3, calculate the total stock, and its expression is s i = Σs ij Step S1.4, calculate the lithium battery production lead date: Based on the inventory situation, obtain the production lead time T of the i-th type of lithium battery i ; The production model established in the said Step S2 with the maximization of lithium battery profit as the goal is: Among them, N types of batteries are produced within a certain time period, and λ i is the selling unit price of the i-th lithium battery; o i is the final production volume of lithium battery i; v ij represents the production cost of the i-th lithium battery in the j-th process; l ij is the storage cost required for the i-th lithium battery in the j-th process; q jm is the consumption of the m-th raw material in the j-th process; w jm is the cost of the m-th raw material in the j-th process; β i is the deferred payment cost; d i is the contractually stipulated delivery volume of the i-th lithium battery.

2. The method according to claim 1, characterized in that, The fitness function in the said Step S3 is: fitness(i,j) = eT ij Among them, eT ij = sT ij + PT ij where eT ij represents the end time of the i-th lithium battery in the j-th process, sT ij is its start time, and PT ij is its total processing time.

3. The method according to claim 2, wherein In the said genetic algorithm, the crossover operation adopts the two-point crossover method.

4. The method according to claim 3, wherein In the said genetic algorithm, the mutation operation adopts the reverse order method.

5. The method according to claim 4, characterized in that, In the said genetic algorithm, the selection operation adopts the roulette wheel selection method.

6. The method according to claim 5, characterized in that, The said method performs production sequencing for different types of battery orders.

7. Application of the method according to any one of claims 1-6 in battery production management.

Images