A fast statistical method for aviation equipment manufacturing process capability data

Abstract

The present invention belongs to the technical field of process capability statistics, and relates to a method for rapid statistics of aviation equipment manufacturing process capability data. Through the interactive integration method of multi-business platform databases, the scope of data calls is narrowed, the operating load of the information management system is reduced, and the system operating efficiency is improved; and batch management means are adopted to implement dynamic management of the database, realize system data validity verification, sample quantity verification and other functions, and ensure data integrity and validity; the present invention satisfies the real-time automatic calculation of data, generates control charts and process capability evaluation and improvement suggestions through abnormality judgment criteria, normality evaluation and calculation models. The present invention can meet the requirements for rapid statistics of aviation product manufacturing process capability data.

Description

Technical Field

[0001] The invention relates to the technical field of process capability statistics, and in particular to a method for quickly counting aviation equipment manufacturing process capability data. Background Art

[0002] In recent years, statistical process control has been widely used to conduct technical analysis of production processes. It can timely discover signs of systematic factors based on feedback information, so that the process can be maintained in a controlled state affected only by random factors. The process capability index is a more convenient and intuitive way to express process fluctuations under the steady state of the statistical process. It is a statistical control tool for measuring process quality. It is currently being applied and promoted in the field of military aviation aircraft. Military aviation aircraft are different from general civilian products. They have a large number of parts and multi-state small batches. They have higher requirements for technical indicators such as product interchangeability, reliability, stability, and fatigue resistance. The continuous and stable operation of key processes can effectively improve manufacturing accuracy and reduce assembly stress. Rapid statistics of manufacturing process capability data are of great significance to improving the stability of the production process, effectively achieving the above technical indicators, and strengthening batch management. Therefore, it is necessary to conduct in-depth research on the statistical technology of process capability data.

[0003] The existing difficulties in the statistical process are: first, existing aviation equipment manufacturing units can generate millions of character data every day, and involve data transmission between the three systems of workshop information collection system K1, information management system K2, and work dispatch management system K3. The data transmission volume is extremely large, which tests the system's carrying capacity; second, data needs to be searched for a set of data sets in a huge database through characteristic information, which is not only slow but also affects calculations and post-processing; third, product characteristic version management, products involve different states, different batches and batch cutting, so it is necessary to judge the validity of product batches, which requires batch control of product characteristics and measured values. Summary of the invention

[0004] The present invention proposes a method for rapid statistical analysis of aviation equipment manufacturing process capability data, which can realize rapid collection, hierarchical and classified statistics of key process capability data and has the characteristics of being fast, stable and convenient for data post-processing.

[0005] A method for quickly calculating aviation equipment manufacturing process capability data comprises the following steps:

[0006] Step 1: Define three databases A, B and C in the system. Database A collects process information, characteristic information, sample quantity and valid batch number.

[0007] Step 2: Collect data from the workshop information collection system K1 at the production site, and push the manufacturing data completed on site to the information management system K2 through the data interface;

[0008] Step 3: Filter invalid data against database A and convert source code, store the converted manufacturing data in database B according to specific data items, and combine databases A and B through the A∪B data set to form database C; wherein the specific data items are in the form of product data information "M" + dispatch data information "O" + manufacturing data information "S", and the form of database B is B = {M1, M3, O, S1, S2}, where M1 represents process information, M3 represents sample quantity information, S1 represents manufacturing information, and S2 represents measured values;

[0009] Step 4: Perform feature screening through database C to determine product information, characteristics, measured values ​​and sample size. Create control charts and distribution charts, determine steady-state and normal distribution, calculate process capability index, and give process capability evaluation and improvement suggestions according to the evaluation model.

[0010] Furthermore, step 1 includes the following sub-steps:

[0011] Step 1-1: Define three databases A, B and C, and read the process information M1 and characteristic information M2 in the digital model into database A;

[0012] Step 1-2: Based on M1 and M2, add the total sample information M3 and the valid batch information M4, determine the four main identification items, define the process that needs to be controlled, and mark it as A = {M1, M2, M3, M4};

[0013] Furthermore, step 2 includes the following sub-steps:

[0014] Step 2-1: Establish the data interface between the workshop information collection system K1 and the network data middle station;

[0015] Step 2-2: The manufacturing records filled out on site, including process information M1, sample quantity information M3, manufacturing data information S, and dispatch data information O, are directly pushed from the workshop information collection system K1 to the management information system K2 in real time through the data interface.

[0016] Furthermore, step 3 includes the following sub-steps:

[0017] Step 3-1: All received data are collated in the management information system K2 according to the specific data item M+O+S format, compared with the process information M1 of the database A, and marked as B={M1, M3, O, S1, S2}, and saved in the database B;

[0018] Step 3-2: Compare the database B with the dispatch management system K3. For the dispatch data information, there exists O=f(M4). Each dispatch data information O can be converted into valid batch number information M4.

[0019] Step 3-3: According to the defined process information M1 and valid batch information M4, check the total number of samples collected with the sample size in database A to verify data integrity;

[0020] Step 3-4: Use the A∪B data set to form complete data information (process information "M1" + characteristic information "M2" + total sample quantity information "M3" + valid batch information "M4" + manufacturing information "S1" + measured value "S2") for database A and database B, and save it to database C, marked C = {M1, M2, M3, M4, S1, S2}.

[0021] Furthermore, step 4 includes the following sub-steps:

[0022] Step 4-1: When performing data analysis and processing, the process information M1 and the effective batch information M4 can be directly retrieved from the database C to determine the product information, characteristics, measured values ​​and sample quantity;

[0023] Step 4-2: Draw a control chart based on the data and combine it with the eight abnormality criteria specified in GB / T 4091-2001 "Conventional Control Chart" to determine whether the process is under steady-state control and whether it has normality;

[0024] Step 4-3: Calculate the process capability index by characteristic type.

[0025] According to the characteristics' own attributes, the characteristics are divided into large-scale characteristics, small-scale characteristics and eye-level characteristics. The three are used to calculate the key process capability index according to different models, but no matter it is large-scale characteristics, small-scale characteristics or eye-level characteristics, it is necessary to calculate the mean value, standard deviation and process capability index Cp of the measured values ​​of the samples.

[0026] Calculate the mean μ of the measured values ​​of the sample:

[0027]

[0028] Where n represents the number of samples, x i represents the i-th sample value;

[0029] Calculate the standard deviation σ:

[0030]

[0031] Cp is calculated by analyzing the manufacturing system without deviation of the intermediate value:

[0032]

[0033] Among them, USL represents the upper limit of tolerance, and LSL represents the lower limit of tolerance;

[0034] When the characteristic is a maximum characteristic, Cpk = Cpu, and Cpu should be calculated:

[0035]

[0036] Among them, Cpk represents the biased process capability index, and Cpu represents the upper biased process capability index;

[0037] When the characteristic is a small characteristic, Cpk = Cpl, and Cpl should be calculated:

[0038]

[0039] Among them, Cpl represents the lower process capability index.

[0040] When the characteristic is a target characteristic, the deviation of the average value from the median value should be considered. To ensure process realization, the product manufacturing baseline is adjusted, and the result of process planning is used as the median value. When the average value deviates from the upper tolerance limit, Cpk = Cpu; when the average value deviates from the lower tolerance limit, Cpk = Cpl.

[0041] Step 4-4: Based on the obtained process capability index, the process capability assessment and improvement suggestions are given from the evaluation dimensions of "Cpk<0.67, 0.67≤Cpk<1.0, 1.0≤Cpk<1.33, 1.33≤Cpk<1.67, Cpk>1.67" and "ε>1.5σ,ε≤1.5σ" according to the calculation model.

[0042] The present invention has the following beneficial effects: through the interactive integration method of multi-business platform databases, the scope of data calls is reduced, the operating load of the information management system is reduced, and the system operating efficiency is improved; and the batch management method is adopted to implement dynamic management of the database, realize the functions of system data validity verification and sample quantity verification, and ensure the integrity and validity of the data; the present invention satisfies the real-time automatic calculation of data, generates control charts and process capability evaluation and improvement suggestions through the judgment criteria, normality evaluation and calculation model. The present invention can meet the requirements for rapid statistics of aviation product manufacturing process capability data. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The present invention is a flow chart of a method for quickly counting aviation equipment manufacturing process capability data.

[0044] Figure 2 Statistical control charts of the measured values ​​of characteristics of a product sample.

[0045] Figure 3 The standard normal distribution plot and the distribution plot of the sample measured values. DETAILED DESCRIPTION

[0046] The technical solution of the present invention is further explained below in conjunction with the accompanying drawings and examples.

[0047] like Figure 1 As shown, a method for quickly calculating the capability data of aviation equipment manufacturing process includes the following steps:

[0048] Step 1: Define three databases A / B / C in the system. Database A collects process information, characteristic information, sample quantity, and valid batch number.

[0049] Step 2: Collect data Z from the workshop information collection system K1 at the production site, and synchronously transfer the manufacturing data completed on site to the data middle platform (it can also be forwarded through the data interface).

[0050] Step 3: Filter invalid data against database A and convert source code, store the converted manufacturing data in database B according to specific data items, and combine databases A and B through the A∪B data set to form database C; wherein the specific data items are in the form of product data information "M" + dispatch data information "O" + manufacturing data information "S", and the form of database B is B = {M1, M3, O, S1, S2}, where M1 represents process information, M3 represents sample quantity information, S1 represents manufacturing information, and S2 represents measured values;

[0051] Step 4: Perform feature screening through database C to determine product information, characteristics, measured values ​​and sample size. Create control charts and distribution charts, determine steady-state and normal distribution, calculate process capability index, and give process capability evaluation and improvement suggestions according to the evaluation model.

[0052] Furthermore, step 1 includes the following sub-steps:

[0053] Step 1-1: Define three databases A / B / C, and read the process information M1 and characteristic information M2 in the digital model into database A;

[0054] Step 1-2: Based on M1 and M2, add the total sample information M3 and the valid batch information M4, determine the four main identification items, define the process that needs to be controlled, and mark it as A = {M1, M2, M3, M4};

[0055] Furthermore, step 2 includes the following sub-steps:

[0056] Step 2-1: Establish the data interface between the workshop information collection system K1 and the network data middle station;

[0057] Step 2-2: The manufacturing records filled out on site, including process information M1, sample quantity information M3, manufacturing data information S, and dispatch data information O, are directly pushed from the workshop information collection system K1 to the management information system K2 in real time through the data interface.

[0058] Furthermore, step 3 includes the following sub-steps:

[0059] Step 3-1: All received data are collated in the management information system K2 according to the specific data item M+O+S format, compared with the process information M1 of the database A, and marked as B={M1, M3, O, S1, S2}, and saved in the database B;

[0060] Step 3-2: Compare the database B with the dispatch management system K3. For the dispatch data information, there exists O=f(M4). Each dispatch data information O can be converted into valid batch number information M4.

[0061] Step 3-3: According to the defined process information M1 and valid batch information M4, check the total number of samples collected with the sample size in database A to verify data integrity;

[0062] Step 3-4: Use the A∪B data set to form complete data information (process information "M1" + characteristic information "M2" + total sample quantity information "M3" + valid batch information "M4" + manufacturing information "S1" + measured value "S2") for database A and database B, and save it to database C, marked C = {M1, M2, M3, M4, S1, S2}.

[0063] Furthermore, step 4 includes the following sub-steps:

[0064] Step 4-1: When performing data analysis, the process information M1 and the effective batch information M4 can be directly retrieved from the database C to determine the product information, characteristics, measured values ​​and sample quantity. Take the sample characteristic data of a selected batch of products as an example (a total of 50), the measured values ​​of all samples are 96, 95.5, 95.5, 96, 96.5, 96, 96.5, 96, 95.5, 95, 95.5, 96, 95.5, 96, 95.5 , 95.5, 95, 95.5, 95, 95.5, 95, 94.5, 95, 94.5, 95, 95.5, 95.5, 95, 95.5, 95, 95.5, 95, 95.5, 95, 95.5, 95, 94.5, 94.5, 94.5, 95, 94.5, 94.5, 95, 94.5, 95, 94.5, 95, 94.5, 95, 94.5, 95, 94.5, 94, 95 ...

[0065] Step 4-2: Draw a control chart using the data, see Figure 2 , combined with the 8 abnormality criteria specified in GB / T 4091-2001 "Conventional Control Chart", it is judged that there is no abnormal data, the process belongs to process steady-state control, and the sample measured value distribution diagram is drawn. The basic form conforms to the normal distribution, see Figure 3 ;

[0066] Step 4-3: According to the characteristics’ own attributes, the characteristics are divided into large-scale characteristics, small-scale characteristics, and eye-level characteristics. The three are used to calculate the key process capability index according to different models, but no matter it is large-scale characteristics, small-scale characteristics, or eye-level characteristics, it is necessary to calculate the mean value, standard deviation, and Cp of the measured values ​​of the samples.

[0067] Calculate the mean μ of the measured values ​​of the sample:

[0068]

[0069] Calculate the standard deviation σ:

[0070]

[0071] Cp is calculated by analyzing the manufacturing system without deviation of the intermediate value:

[0072]

[0073] This characteristic is a target characteristic. The deviation of the average value from the median value should be considered. To ensure process realization, the product manufacturing baseline should be adjusted appropriately, with the result of process planning as the median value. When the average value deviates from the upper tolerance limit (ε>0), Cpk=Cpu; when the average value deviates from the lower tolerance limit (ε<0), Cpk=Cpl.

[0074] The average value of the selected sample is μ=95.2, so it is an upward deviation, so:

[0075]

[0076] Step 4-4: Based on the obtained process capability index, the process capability evaluation and improvement suggestions are given from the evaluation dimensions of "Cpk<0.67, 0.67≤Cpk<1.0, 1.0≤Cpk<1.33, 1.33≤Cpk<1.67, Cpk>1.67" and "ε>1.5σ,ε≤1.5σ" according to the calculation model. After calculation, Cpk=1.39, 1.33≤Cpk<1.67, and ε≤1.5σ of the product sample, which belongs to the process capability index evaluation level II, indicates that the process capability is sufficient, the process operation is stable during the sample period, and the technical management capability is good, which should be maintained.

[0077] The present invention is based on the use of a multi-business platform database interaction integration method, combined with a process capability calculation model, to reduce the data call range on the existing system to 0.72% of the total data, and realize the automatic verification of data validity, sample quantity and other functions, especially for the data interaction preprocessing between different business systems, which has a good reference. The method is highly operable, fast and flexible in data processing, and can be widely used in scenarios where independent business management systems are used, product batches are small, multiple states are present, and there is a need for rapid statistical process capability data.

[0078] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A method for rapid statistical analysis of aviation equipment manufacturing process capability data, characterized in that: The steps include: Step 1: Define three databases A, B and C in the system. Database A collects process information, characteristic information, sample quantity and valid batch number. Step 1-1: Define three databases A, B and C, and read the process information M1 and characteristic information M2 in the digital model into database A; Step 1-2: Based on M1 and M2, add the total sample information M3 and the valid batch information M4, determine the four main identification items, define the process that needs to be controlled, and mark it as A = {M1, M2, M3, M4}; Step 2: Collect data from the workshop information collection system K1 at the production site, and push the manufacturing data completed on site to the information management system K2 through the data interface; Step 2-1: Establish the data interface between the workshop information collection system K1 and the network data middle station; Step 2-2: The manufacturing records filled in on-site, including process information M1, sample quantity information M3, manufacturing data information S, and dispatch data information O, are directly pushed from the workshop information collection system K1 to the management information system K2 in real time through the data interface; Step 3: Filter invalid data against database A and convert source code, store the converted manufacturing data in database B according to specific data items, and combine databases A and B through the A∪B data set to form database C; where the specific data items are in the form of product data information "M" + dispatch data information "O" + manufacturing data information "S", and the form of database B is B = {M1, M3, O, S1, S2}, where M1 represents process information, M3 represents sample quantity information, S1 represents manufacturing information, and S2 represents measured values; Step 3-1: All received data are collated in the management information system K2 according to the specific data item M+O+S format, compared with the process information M1 of the database A, and marked as B={M1, M3, O, S1, S2}, and saved in the database B; Step 3-2: Compare the database B with the dispatch management system K3. For the dispatch data information, there exists O=f(M4). Each dispatch data information O can be converted into valid batch number information M4. Step 3-3: According to the defined process information M1 and valid batch information M4, check the total number of samples collected with the sample size in database A to verify data integrity; Step 3-4: Database A and database B are combined through the A∪B data set to form complete data information, namely process information "M1" + characteristic information "M2" + total sample information "M3" + valid batch information "M4" + manufacturing information "S1" + measured value "S2", and saved to database C, marked C = {M1, M2, M3, M4, S1, S2}; Step 4: Perform feature screening through database C to determine product information, characteristics, measured values ​​and sample size; and form control charts and distribution charts to determine steady-state and normal distribution and calculate process capability index, and give process capability evaluation and improvement suggestions according to the evaluation model.

2. The method for rapid statistical analysis of aviation equipment manufacturing process capability data according to claim 1, characterized in that: The step 4 includes the following sub-steps: Step 4-1: When performing data analysis and processing, directly retrieve the process information M1 and the valid batch information M4 from the database C to determine the product information, characteristics, measured values ​​and sample quantity; Step 4-2: Draw a control chart based on the data and combine it with the eight exception criteria specified in GB / T 4091-2001 "Conventional Control Chart" to determine whether the process is under steady-state control and whether it is normal; Step 4-3: Calculate process capability index by characteristic type According to the characteristics' own attributes, the characteristics are divided into large-scale characteristics, small-scale characteristics and large-scale characteristics. The three are used to calculate the key process capability index according to different models. However, no matter it is large-scale characteristics, small-scale characteristics or large-scale characteristics, it is necessary to calculate the mean value, standard deviation and process capability index Cp of the measured values ​​of the samples. Calculate the mean μ of the measured values ​​of the sample: Where n represents the number of samples, x i represents the i-th sample value; Calculate the standard deviation σ: Cp is calculated by analyzing the manufacturing system without deviation of the intermediate value: Among them, USL represents the upper limit of tolerance, and LSL represents the lower limit of tolerance; When the characteristic is a maximum characteristic, Cpk = Cpu, calculate Cpu: Among them, Cpk represents the biased process capability index, and Cpu represents the upper biased process capability index; When the characteristic is a small characteristic, Cpk = Cpl, calculate Cpl: Among them, Cpl represents the lower process capability index; When the characteristic is a target characteristic, the deviation of the average value from the median value is considered. To ensure process realization, the product manufacturing baseline is adjusted, and the result of process planning is used as the median value. When the average value deviates from the upper tolerance limit, Cpk = Cpu; when the average value deviates from the lower tolerance limit, Cpk = Cpl. Step 4-4: Based on the obtained process capability index, the process capability evaluation and improvement suggestions are given from the evaluation dimensions of "Cpk<0.67, 0.67≤Cpk<1.0, 1.0≤Cpk<1.33, 1.33≤Cpk<1.67, Cpk>1.67" and "ε>1.5σ,ε≤1.5σ" according to the calculation model.

Images