Preparation for 2022 Supervision Engineer: Day 4

2022-06-24 0 By

Section 1 Statistical analysis of engineering quality i. Basic principles and methods of engineering quality statistics and sampling inspection (1) Characteristic values of quality data (memory: two numbers, two difference and one coefficient) The characteristic values of the central trend: arithmetic mean and median;The eigenvalues of the discrete trend: poor, standard deviation, variation coefficient, the smaller the standard deviation or coefficient of variation, high degree of concentration distribution, the discrete degree is small, the average for the representation of the overall good quality data (2) the characteristics of the distribution characteristics of quality data, the volatility of individual value b, overall distribution regularity (discrete) concentration, 2.The regular distribution of quality data is the most important, common and widely used.3. Reason for fluctuation of quality data Accidental cause (normal) : small change of 4M1E factor;Random features are unavoidable, difficult to measure and control, or not economically worth eliminating.They are abundant, but have little effect on quality.System cause (abnormality) : 4M1E factor changed greatly;Workers do not comply with the operation procedures, mechanical equipment failure or excessive wear, raw material quality specifications significant differences, not eliminated in time, the production process is abnormal.(3) Reasons for sampling inspection and inspection lotDamage, time, money and experience) 1) destructive inspection, can’t take the book inspection way 2) for the entire quantity sometimes need to spend a large cost, on the economy does not necessarily buy 3) inspection needs time, the book inspection on time sometimes don’t allow 4) even if the full inspection, does not necessarily 100% 5) sampling inspection qualified sampling inspection personnel at the mercy of the subjective intention,The odds are the same.The sample is evenly distributed and fully representative.Can be used for quality control of destructive or manufacturing processes.(4) simple random sampling method (nickname: pure random, completely random) system (nickname: mechanical random) random stratified random (exception: stratified proportional sampling) multi-stage sampling (5) the classification of sampling inspection and sampling plan 1, classification: measurement model: a decimal point, such as: weight, strength and geometry size, elevation, displacement of counting type: an integer, such as:Bad number, unqualified emitting, defects of solder joints for a sampling parameters, such as: N, N, C (three parameters) judgment method: sample N check out unqualified play: d if d < = C, then determine the eligible if d > C, then determine the unqualified twice sampling parameters: N, n1, n2, C1 and C2 (five parameters) judgment method:Number of unqualified products tested by the first sample sampling n1: d1 Number of unqualified products tested by the second sample sampling n2: d2 If D1 >C2, it will be judged as unqualified. If C1 < D1 <= C2, continue to judge:When d1+ D2 <=C2, it is judged qualified. When D1 + D2 > C2, it is judged unqualified. 2.False error, unqualified batch is judged as qualified batch, probability is β, user risk control project, α, β should not exceed 5%;General items, α should not exceed 5%, β should not exceed 10% ii. Statistical analysis method of engineering quality (I) questionnaire method (statistical investigation analysis) Using tools: using special statistical tables: collection, sorting and rough analysis (II) stratified method (classification) method:1. Observation and analysis of the permutation chart 1) Each histogram represents a quality problem and influencing factors, and the degree of influence is proportional to the height.2) Determine the primary and secondary factors 0–80% major factors 80%–90% secondary factors 90%–100% general factors 2, the application of the chart according to the content of the unqualified point classification —– analysis of the weak links of quality problems according to the classification of production operations —–Find out the key process of producing the most nonconforming products according to the classification of production teams or units —– analyze and compare the arrangement chart before and after the technical and management level of each unit to take improvement measures —–Analysis measures are effective for cost analysis, safety analysis, etc. (four) causal analysis (characteristic factor diagram, branch diagram or fish bone diagram) analysis of quality problems (results) and the relationship between the cause of the effective tool should be paid attention to when drawing and using causal analysis diagram:Histogram method used to describe the state of mass distribution of an analysis method, also known as the mass distribution map method 1, histogram method of use:1) Understand the fluctuation of product quality and master the distribution law of quality characteristics; 2) estimate the overall rate of nonconforming product in the construction and production process and evaluate the process capability through the calculation of quality data characteristic values;Group spacing improperly determined left (right) gentle slope type — cause — Island type caused by too strict upper or lower limit control in operation — cause — change of raw materials/temporary shift by others — cause — two different methods or two groups of workers production,3. Compare the histogram with the quality standard to judge that the actual production process capacity data distribution is lower than the lower limit (the left side is the lower limit), which is prone to unqualified.In terms of management, it is necessary to improve the distribution width boundary of the overall capacity data to reach the upper and lower limits of the quality standard. The quality capacity is in a critical state and is prone to unqualified, so it is necessary to analyze the causes and take measures. The distribution of the quality data is in the middle and the boundary has a large distance from the upper and lower limits of the quality standard, indicating that the quality capacity is relatively large.No economic data distribution have been beyond the upper and lower bounds of quality standards, the data shows that the production process quality is unqualified, to analysis the reason, take measures to rectify (6) control chart (management chart) use of control chart to distinguish the quality fluctuation reasons, determine whether the production process in a stable state method referred to as the control method of figure 1, 1) process analysis, the purpose of the control chartThat is to analyze whether the production process is stable 2) process control, control the quality state of the production process 2) observation and analysis of the control diagram Purpose of drawing the control diagram: analyze and judge whether the production process is in a stable state.1) Ideas almost all fall within the control limit a, more than 25 consecutive points within the control limit.B) Only 1 out of 35 consecutive points is beyond the control limit.C. No more than 2 out of 100 consecutive points exceed the control limit.(100-2) 2) There are no defects in the arrangement of ideas within the control boundary. Defects and abnormal phenomena occur in the arrangement of ideas: A, the occurrence of “chain”, that is, seven consecutive points on the same side (memory: 5 attention, 6 investigation and 7 processing) B, the occurrence of many times on the same side, that is:11/10, 14/12, 17/14, 20/16c, the occurrence of trend or tendency, that is, the continuous rise or decline of seven consecutive points or more d, the occurrence of periodic changes, similar to the normal distribution E, the occurrence of point arrangement close to the control line, that is, 3/2, 7/3,10/4 (7) Correlation diagram method to show the relationship between two kinds of quality data 1, the relationship between quality characteristics and influencing factors 2, the relationship between quality characteristics and quality characteristics 3, the relationship between influencing factors and influencing factors