Production line operation and maintenance management system based on big data

Abstract

The invention discloses a production line operation and maintenance management system based on big data, and relates to the technical field of production line management, and the system comprises a production and storage module which is used for calculating an optimal production and storage plan; the optimal output module is used for calculating the optimal output of each production line; the production plan making module is used for making a work plan of each production line; the recording and feedback module is used for establishing a production log list and recording and feeding back actual production conditions; and the cloud storage, arrangement and analysis module is used for establishing a cloud database platform and storing, arranging and analyzing the production logs. The method has the advantages that the daily average optimal production output of each production line is calculated by analyzing the customer demand quantity and the production output of each production line, the problem is found in time and the production mode is adjusted by comparing the actual value and the predicted value of the daily average production output, more elastic mechanisms are given to the production line, the efficiency is improved, and the production cost is reduced. And resource waste is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of production line management, and in particular to a production line operation and maintenance management system based on big data. Background Art

[0002] Production line management plays a vital role in modern manufacturing. Reasonable production processes and task allocation can effectively improve production efficiency and reduce the problem of overcapacity. In addition, in the face of changes in market demand, companies need to quickly adjust production plans, optimize production layout and resource allocation. It will also be particularly important to study production line operation and maintenance management systems with high flexibility and adaptability.

[0003] The existing production line management system mainly focuses on fault prevention and production line detection on production line equipment. It lacks analysis of the supply and demand balance between customer demand and production. In the actual production process, it relies on the subjective judgment of staff, resulting in chaotic production methods, overcapacity and substandard production, causing waste of resources and low production efficiency, which is not conducive to the long-term development of the enterprise. Summary of the invention

[0004] In order to solve the above technical problems, a production line operation and maintenance management system based on big data is provided. This technical solution solves the problem proposed in the above background technology that there is a lack of analysis of the supply and demand balance between the customer demand side and the production side, and that the actual production process relies on the subjective judgment of the staff, resulting in chaotic production methods, overcapacity and substandard production in actual production.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A production line operation and maintenance management system based on big data, comprising:

[0007] A production and storage module, which is used to calculate the optimal production and storage plan according to customer needs;

[0008] An optimal production volume module, which is used to calculate the optimal production volume of each production line based on big data;

[0009] A production plan making module, which is used to make a work plan for each production line according to the optimal production and storage plan and the optimal quantity of the production line;

[0010] A recording and feedback module, which is used to establish a production log list, record and provide feedback on the actual production situation;

[0011] The cloud storage, sorting and analysis module is used to establish a cloud database platform to store, sort and analyze production logs.

[0012] Preferably, the production and storage module specifically includes:

[0013] Demand analysis unit, the demand analysis unit is used to establish a phased list according to customer demand and production and storage conditions, where the number of stages is n, and the customer demand in the i-th stage is N i To express, the staged list includes: the cost of producing a single product unit, the cost of storing a single product unit, the number of stages, and the customer demand at each stage;

[0014] A production cost optimal index unit, which is used to construct an index function for generating costs at each stage and screen the optimal index function based on production and storage information;

[0015] A state transfer unit, wherein the state transfer unit is used to establish a state transfer equation according to the inventory, production and customer demand in each stage;

[0016] The optimal production and storage planning unit is used to screen out the optimal production and storage plan through calculation, that is, the optimal S k , X k and the value of the minimum cost.

[0017] Preferably, the indicator function for generating costs at each stage is:

[0018]

[0019] Among them, g(S k X k ) is the cost of the kth stage, a is the cost of producing a single product unit, b is the cost of storing a single product unit, S k is the state variable, i.e. the inventory at the beginning of the kth stage, X k is the decision variable, i.e., the production volume in the kth stage;

[0020] The optimal indicator function is:

[0021] f k (S k )=min{g(S k X k )+f k+1 (S k+1 )} X k ∈X k (S k )

[0022] Among them, f k (S k ) refers to the optimal solution of the indicator function from the kth stage to the nth stage, g(S k X k ) is the cost of the kth stage, f k+1 (S k+1 ) is the cost of the k+1th stage;

[0023] The state transfer equation is:

[0024] S k+1 =S k +X k -d k

[0025] Among them, S k+1 is the inventory at stage k, S k is the inventory at the beginning of the kth stage, X k is the production volume of the kth stage, d k is the customer demand in the kth stage.

[0026] Preferably, the optimal production volume module specifically includes:

[0027] A linear regression unit, wherein the linear regression unit is used to construct a linear relationship between production volume, production efficiency, accident volume and delay time based on big data;

[0028] The error processing unit is used to construct an error relationship between the true value and the predicted value, where the error is subject to a mean of 0 and a variance of σ 2 Gaussian distribution of

[0029] An objective function unit, wherein the objective function unit is used to construct an objective function about a parameter based on a Gaussian distribution function;

[0030] The optimal production volume prediction unit is used to find the optimal solution between the real value and the predicted value based on the gradient descent algorithm, and then predict the optimal production volume of each production line under the influence of several characteristic items such as production efficiency, accident volume and delay time according to the optimal solution.

[0031] Preferably, the linear expressions of production volume, production efficiency, accident volume and delay time are:

[0032]

[0033] Among them, y is the predicted value of the production volume of each stage production line, α i is the parameter of the i-th feature item, x i is the i-th feature item;

[0034] The error relationship between the true value and the predicted value is:

[0035] y ε =|y z -y|

[0036] Among them, y ε is the error between the true value and the predicted value, y z is the actual value recorded in the big data, and y is the predicted value of the production volume of the production line at each stage;

[0037] The objective function of the parameters is:

[0038]

[0039] Among them, f α is the relationship between the predicted value and the actual value, y z is the actual value recorded in the big data, y is the predicted value of the production volume of the production line at each stage, and m is the number of samples of the relationship between the predicted value and the actual value;

[0040] The expression of the gradient descent algorithm is:

[0041] α t =α t-1 -β*h t

[0042] Among them, α t is the parameter value of a certain characteristic item at time t, α t-1 is the parameter value of a feature item at time t-1, β is the descent step length in the gradient descent process, and h t is the gradient of the objective function at time t, where h t It is calculated by taking the partial derivative of the objective function.

[0043] Preferably, formulating a work plan for each production line according to the optimal production and storage plan and the optimal quantity of the production line specifically includes:

[0044] According to the above-mentioned optimal production and storage plan and the calculation results of the optimal production volume of each production line, determine the task volume of each production line in the staged work;

[0045] The phased task volume is averaged to the daily production volume and set as the initial value of the average daily workload of each production line.

[0046] Preferably, the establishment of a production log list to record and provide feedback on the actual production situation specifically includes:

[0047] Set up a production log list submission point on each production line and fill in the relevant production information of the production line every day;

[0048] The relevant production information fed back is uploaded to the database through the cloud, wherein the production log list includes: daily production volume, working hours, number of abnormal accidents, delay time, production efficiency, person in charge, and date.

[0049] Preferably, the establishment of a cloud database platform to store, organize and analyze production logs specifically includes:

[0050] Build a cloud database platform to store, organize and analyze historical and updated production log data;

[0051] Record, classify and count the production logs submitted daily by each production line according to their contents;

[0052] Compare actual daily performance of each production line with forecast performance:

[0053] When the actual daily average output is less than the predicted output, it means that the existing allocation method is unable to complete the task, and the subsequent output needs to be redistributed by averaging the remaining output;

[0054] When the actual daily average output is equal to the predicted output, it means that the daily output meets the standard and no redistribution is required;

[0055] When the actual average daily output is greater than the predicted amount, it means that the existing allocation method can complete the task, and the subsequent output will be redistributed according to normal production according to the initial output or by averaging the remaining output.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The present invention proposes a production line operation and maintenance management system based on big data. By analyzing the customer demand and production volume at each stage, the supply and demand balance between the client and the production end is found, the optimal production volume of each production line is calculated, and on this basis, the workload of each production line and the average daily production volume of each production line are calculated. By comparing the actual value of the average daily production volume with the predicted value, problems are discovered and the production mode is adjusted in time, giving the production line more flexibility mechanisms, improving efficiency and reducing waste of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is a structural block diagram of a production line operation and maintenance management system based on big data proposed by the present invention;

[0059] Figure 2 Screening out the optimal production and storage plan flow chart for the present invention;

[0060] Figure 3A flowchart for predicting the optimal production volume of each production line based on big data according to the present invention;

[0061] Figure 4 This is a flow chart of the present invention for reallocating subsequent production according to the daily production status of each production line displayed on the cloud database platform. DETAILED DESCRIPTION

[0062] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0063] Reference Figure 1 As shown, a production line operation and maintenance management system based on big data includes:

[0064] A production and storage module, which is used to calculate the optimal production and storage plan according to customer needs;

[0065] An optimal production volume module, which is used to calculate the optimal production volume of each production line based on big data;

[0066] A production plan making module, which is used to make a work plan for each production line according to the optimal production and storage plan and the optimal quantity of the production line;

[0067] A recording and feedback module, which is used to establish a production log list, record and provide feedback on the actual production situation;

[0068] The cloud storage, sorting and analysis module is used to establish a cloud database platform to store, sort and analyze production logs.

[0069] It can be explained that this solution is based on the analysis of the three aspects of client, production and data storage in combination with actual conditions. First, according to the customer needs at each stage and in accordance with the principle of maximizing benefits, the optimal production and storage plan is screened out, where the screened production volume is the total production volume required at each stage; secondly, according to the actual production situation of each production line recorded in big data, the optimal quantity that each production line can produce is screened out through the gradient descent algorithm, and then the specific daily work tasks can be accurately arranged for each production line, and the daily work situation can be fed back to the cloud database through the production log list for recording and analysis, which is convenient for understanding the actual production situation and making adjustments, and also provides new data for the next production calculation.

[0070] Reference Figure 2 As shown, the screening out of the optimal production and storage plan specifically includes:

[0071] Demand analysis unit, the demand analysis unit is used to establish a phased list according to customer demand and production and storage conditions, where the number of stages is n, and the customer demand in the i-th stage is N i To express, the staged list includes: the cost of producing a single product unit, the cost of storing a single product unit, the number of stages, and the customer demand at each stage;

[0072] A production cost optimal index unit, which is used to construct an index function for generating costs at each stage and screen the optimal index function based on production and storage information;

[0073] A state transfer unit, wherein the state transfer unit is used to establish a state transfer equation according to the inventory, production and customer demand in each stage;

[0074] The optimal production and storage planning unit is used to screen out the optimal production and storage plan through calculation, that is, the optimal S k , X k and the value of the minimum cost.

[0075] The indicator function for the cost generated in each stage is:

[0076]

[0077] Among them, g(S k X k ) is the cost of the kth stage, a is the cost of producing a single product unit, b is the cost of storing a single product unit, S k is the state variable, i.e. the inventory at the beginning of the kth stage, X k is the decision variable, i.e., the production volume in the kth stage;

[0078] The optimal indicator function is:

[0079] f k (S k )=min{g(S k X k )+f k+1 (S k+1 )}X k ∈X k (S k )

[0080] Among them, f k (S k ) refers to the optimal solution of the indicator function from the kth stage to the nth stage, g(S k X k ) is the cost of the kth stage, f k+1(S k+1 ) is the cost of the k+1th stage;

[0081] The state transfer equation is:

[0082] S k+1 =S k +X k -d k

[0083] Among them, S k+1 is the inventory at stage k, S k is the inventory at the beginning of the kth stage, X k is the production volume of the kth stage, d k is the customer demand in the kth stage.

[0084] It can be explained that the optimal production and storage plan is to ensure the balance between supply and demand. The production volume of this type of product is exactly equal to the customer demand, that is, the inventory volume at the beginning and end of production is 0. Then according to this condition, the boundary conditions for calculating the optimal solution of the indicator function can be obtained, and the solution method of the function is the reverse method, wherein the reverse method is to analyze the necessary conditions of each conclusion, start, and reverse step by step until the conditions given in the question just meet the requirements. When using the reverse method, the more stages there are, the more complicated the calculation is. Therefore, in order to solve the complexity of the calculation and reduce the load of the system, the number of stages should be controlled when analyzing the needs of staged customers.

[0085] Reference Figure 3 As shown, the method of predicting the optimal production volume of each production line based on big data specifically includes:

[0086] A linear regression unit, wherein the linear regression unit is used to construct a linear relationship between production volume, production efficiency, accident volume and delay time based on big data;

[0087] The error processing unit is used to construct an error relationship between the true value and the predicted value, where the error is subject to a mean of 0 and a variance of σ 2 Gaussian distribution of

[0088] An objective function unit, wherein the objective function unit is used to construct an objective function about a parameter based on a Gaussian distribution function;

[0089] The optimal production volume prediction unit is used to find the optimal solution between the real value and the predicted value based on the gradient descent algorithm, and then predict the optimal production volume of each production line under the influence of several characteristic items such as production efficiency, accident volume and delay time according to the optimal solution.

[0090] The linear expressions of production volume, production efficiency, accident volume and delay time are as follows:

[0091]

[0092] Among them, y is the predicted value of the production volume of each stage production line, α i is the parameter of the i-th feature item, x i is the i-th feature item;

[0093] The error relationship between the true value and the predicted value is:

[0094] y ε =|y z -y|

[0095] Among them, y ε is the error between the true value and the predicted value, y z is the actual value recorded in the big data, and y is the predicted value of the production volume of the production line at each stage;

[0096] The objective function of the parameters is:

[0097]

[0098] Among them, f α is the relationship between the predicted value and the actual value, y z is the actual value recorded in the big data, y is the predicted value of the production volume of the production line at each stage, and m is the number of samples of the relationship between the predicted value and the actual value;

[0099] The expression of the gradient descent algorithm is:

[0100] α t =α t-1 -β*h t

[0101] Among them, α t is the parameter value of a certain characteristic item at time t, α t-1 is the parameter value of a feature item at time t-1, β is the descent step length in the gradient descent process, and h t is the gradient of the objective function at time t, where h t It is calculated by taking the partial derivative of the objective function.

[0102] It can be explained that in the actual production process, the production volume will be affected by production efficiency, accidents and delay time, resulting in an error between the actual production and the predicted production. Therefore, this scheme analyzes each influencing characteristic item and establishes a linear expression based on the characteristic item: y = α0 + α1x1 + α2x2 + α3x3. However, in order to facilitate subsequent calculations, by multiplying x0 at α0 and x0 = 1, we construct Expression, secondly, since the gradient descent algorithm is a commonly used algorithm in the process of machine learning, the purpose is to find the minimum value by iterating parameters, and the objective function is established by removing the influence of irrelevant terms of constants and intercepting the part of the relevant parameters after taking the logarithm of the Gaussian distribution expression, so as to meet the gradient descent algorithm and meet the objective function. The minimum value is found by the gradient descent algorithm, which is the optimal solution; the specific logarithm of the Gaussian distribution expression is: It can be simplified to: In order to obtain more accurate results, the results between the predicted value and the actual value are accumulated m times.

[0103] Reference Figure 4 As shown, the redistribution of subsequent production according to the daily production status of each production line displayed by the cloud database platform specifically includes:

[0104] Build a cloud database platform to store, organize and analyze historical and updated production log data;

[0105] Record, classify and count the production logs submitted daily by each production line according to their contents;

[0106] Compare actual daily performance of each production line with forecast performance:

[0107] When the actual daily average output is less than the predicted output, it means that the existing allocation method is unable to complete the task, and the subsequent output needs to be redistributed by averaging the remaining output;

[0108] When the actual daily average output is equal to the predicted output, it means that the daily output meets the standard and no redistribution is required;

[0109] When the actual average daily output is greater than the predicted amount, it means that the existing allocation method can complete the task, and the subsequent output will be redistributed according to normal production according to the initial output or by averaging the remaining output.

[0110] It can be explained that the cloud database platform includes two functions. One part is to record the production log list submitted daily by each production line, and to classify and count the contents; the other part is to obtain the predicted value of daily production volume of each production line through system calculation, and to intuitively display the comparison between actual value and predicted value in the form of a chart, so as to facilitate the later adjustment of the production line.

[0111] To sum up, the advantages of the present invention are: by analyzing the customer demand and production volume at each stage, the supply and demand balance between the client and the production end is found, the optimal production volume of each production line is calculated, and on this basis, the workload of each production line is calculated. The average daily production volume of each production line is calculated, and by comparing the actual value of the average daily production volume with the predicted value, problems are discovered and the production mode is adjusted in time, giving the production line more flexibility mechanisms, improving efficiency and reducing waste of resources.

[0112] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A production line operation and maintenance management system based on big data, characterized in that: include: A production and storage module, which is used to calculate the optimal production and storage plan according to customer needs; An optimal production volume module, which is used to calculate the optimal production volume of each production line based on big data; A production plan making module, which is used to make a work plan for each production line according to the optimal production and storage plan and the optimal quantity of the production line; A recording and feedback module, which is used to establish a production log list, record and provide feedback on the actual production situation; The cloud storage, sorting and analysis module is used to establish a cloud database platform to store, sort and analyze production logs.

2. The production line operation and maintenance management system based on big data according to claim 1, characterized in that: The production and storage module specifically includes: Demand analysis unit, the demand analysis unit is used to establish a phased list according to customer demand and production and storage conditions, where the number of stages is n, and the customer demand in the i-th stage is N j To express, the staged list includes: the cost of producing a single product unit, the cost of storing a single product unit, the number of stages, and the customer demand at each stage; A production cost optimal index unit, which is used to construct an index function for generating costs at each stage and screen the optimal index function based on production and storage information; A state transfer unit, wherein the state transfer unit is used to establish a state transfer equation according to the inventory, production and customer demand in each stage; The optimal production and storage planning unit is used to screen out the optimal production and storage plan through calculation, that is, the optimal S k , X k and the value of the minimum cost.

3. The production line operation and maintenance management system based on big data according to claim 2 is characterized in that: The indicator function for the cost generated in each stage is: Among them, g(S k X k ) is the cost of the kth stage, a is the cost of producing a single product unit, b is the cost of storing a single product unit, S k is the state variable, i.e. the inventory at the beginning of the kth stage, X k is the decision variable, i.e., the production volume in the kth stage; The optimal indicator function is: f k (S k )=min{g(S k X k )+f k+1 (S k+1 )} X k ∈X k (S k ) Among them, f k (S k ) refers to the optimal solution of the indicator function from the kth stage to the nth stage, g(S k X k ) is the cost of the kth stage, f k+1 (S k+1 ) is the cost of the k+1th stage; The state transfer equation is: S k+1 =S k +X k -d k Among them, S k+1 is the inventory at stage k, S k is the inventory at the beginning of the kth stage, X k is the production volume of the kth stage, d k is the customer demand in the kth stage.

4. The production line operation and maintenance management system based on big data according to claim 3 is characterized in that: The optimal production volume module specifically includes: A linear regression unit, wherein the linear regression unit is used to construct a linear relationship between production volume, production efficiency, accident volume and delay time based on big data; The error processing unit is used to construct an error relationship between the true value and the predicted value, where the error is subject to a mean of 0 and a variance of σ 2 Gaussian distribution of An objective function unit, wherein the objective function unit is used to construct an objective function about a parameter based on a Gaussian distribution function; The optimal production volume prediction unit is used to find the optimal solution between the real value and the predicted value based on the gradient descent algorithm, and then predict the optimal production volume of each production line under the influence of several characteristic items such as production efficiency, accident volume and delay time according to the optimal solution.

5. The production line operation and maintenance management system based on big data according to claim 4 is characterized in that: The linear expressions of production volume, production efficiency, accident volume and delay time are as follows: Among them, y is the predicted value of the production volume of each stage production line, α i is the parameter of the i-th feature item, x i is the i-th feature item; The error relationship between the true value and the predicted value is: and ε =|and z -and| Among them, y ε is the error between the true value and the predicted value, y z is the actual value recorded in the big data, and y is the predicted value of the production volume of the production line at each stage; The objective function of the parameters is: Among them, f α is the relationship between the predicted value and the actual value, y z is the actual value recorded in the big data, y is the predicted value of the production volume of the production line at each stage, and m is the number of samples of the relationship between the predicted value and the actual value; The expression of the gradient descent algorithm is: α t ＝α t-1 -β*h t Among them, α t is the parameter value of a certain characteristic item at time t, α t-1 is the parameter value of a feature item at time t-1, β is the descent step length in the gradient descent process, and h t is the gradient of the objective function at time t, where h t It is calculated by taking the partial derivative of the objective function.

6. The production line operation and maintenance management system based on big data according to claim 5, characterized in that: The work plan for each production line is formulated based on the optimal production and storage plan and the optimal quantity of the production line, and specifically includes: According to the above-mentioned optimal production and storage plan and the calculation results of the optimal production volume of each production line, determine the task volume of each production line in the staged work; The phased task volume is averaged to the daily production volume and set as the initial value of the average daily workload of each production line.

7. The production line operation and maintenance management system based on big data according to claim 6 is characterized in that: The establishment of a production log list to record and provide feedback on the actual production situation specifically includes: Set up a production log list submission point on each production line and fill in the relevant production information of the production line every day; The relevant production information fed back is uploaded to the database through the cloud, wherein the production log list includes: daily production volume, working hours, number of abnormal accidents, delay time, production efficiency, person in charge, and date.

8. The production line operation and maintenance management system based on big data according to claim 7, characterized in that: The establishment of a cloud database platform to store, organize and analyze production logs specifically includes: Build a cloud database platform to store, organize and analyze historical and updated production log data; Record, classify and count the production logs submitted daily by each production line according to their contents; Compare actual daily performance of each production line with forecast performance: When the actual daily average output is less than the predicted output, it means that the existing allocation method is unable to complete the task, and the subsequent output needs to be redistributed by averaging the remaining output; When the actual daily average output is equal to the predicted output, it means that the daily output meets the standard and no redistribution is required; When the actual average daily output is greater than the predicted amount, it means that the existing allocation method can complete the task, and the subsequent output will be redistributed according to normal production according to the initial output or by averaging the remaining output.

Images