Dissertation is on the receipt of scientific degree of candidate of technical sciences, specialty 05.01.02 – standardization, certification and metrology. The National Tech-nical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute", MES Ukraine, Kiev, 2019.
The thesis is aimed at addressing the issues of product quality assurance in a low-power enterprise by applying a general approach to assessing the breakdown of the technological process and the formation of criteria for its early detection, the development of statistical methods for ensuring compliance of the finished product.
The work consists in theoretical substantiation, formation and further develop-ment of methods for ensuring the quality of products at the stages of its life cycle. The paper substantiates the concept of a low-power enterprise in order to focus on the power of the enterprise, which does not have a correlated connection with the number of staff, and characterizes its special feature. The technological process of such enter-prises is periodic, includes non-automated operations, and production is small and multicomponent, with the same equipment and means of measurement being used. All this complicates the conduct of control operations and limits them in time.
It is shown the importance of maintaining the stability of production to ensure the quality of products. Taking into account the particularity of a small enterprise, which consists in the frequent change in the product range, its small volume and the lack of a significant amount of sample data, it is essential to identify the process of decomposition in the early stages. For operational control of the technological pro-cess, it is necessary to justify the choice of intervals between the control samples. For this purpose, the sensitivity of different types of cards to the technological process is analyzed. It was established that the recommendation of normative documents regard-ing 4-5 observations in control samples is valid only for s-cards. For medium-value maps, the sensitivity varies by 20% depending on the number of observations. An analysis of the sequence of events, which is displayed by finding the control points in certain areas on the map of average values, is determined and the probability of such a complex event is determined. The efficiency index is introduced, which is the criteri-on for early detection of the process decomposition, and the method of combining el-ementary events has been developed, in which additional criteria were found accord-ing to the binomial model.
When evaluating compliance with fixed limit values, additional limit values are introduced in their vicinity. When the primary result hits the middle of the region, bounded by additional limit s, they pass to a consistent procedure for detecting the ra-tio of the current result and the additional limit values. At the same time, the probabil-ity of making false decisions decreases with insignificant increase in the volume of control and measuring operations.
It is shown that the characteristic feature of low-power enterprises is a wide range of products, each of which has a "own" tolerance on a controlled indicator for the same control equipment. Under these conditions, the method of adaptive control limit values is proposed, where each successive step of conformity assessment leads to an increase in the index of measurement capabilities and, in turn, expands the con-trol limits with which the average value, calculated on all results obtained on all the steps. This method is distinguished by the possibility of gradually increasing the ac-ceptance limits within the tolerance field to the controlled parameter for objects with different lengths of the receiving interval, the correspondence of which has not been established in the previous stages. This method, in contrast to the method with the in-troduction of guard band, allows you to exclude manufacturer losses and provide a given value of the probability of compliance with norms.
The method of planning (designing) quality control of products was developed and tested in the production conditions of confectionery production; practical recom-mendations for its application are given. The introduction of the methodology has made it possible to reduce regulatory losses of raw materials in the production of products and the number of returns from the retail network due to reasons of inade-quate quality by 2-3 percentage points.
Key words: product quality, technological process, control cards, statistical control, criteria, stability and precision of the process, conformity assessment of products, uncertainty of measurement results, adaptive sequential algorithm.
