In engineering practice, these uncertainties are usually divided into the following types:
Randomness is the uncertainty factor that designers first consider. Randomness refers to the uncertainty of a certain outcome of an event, that is, the various results of something under the same conditions. This is due to the insufficient conditions, which prevents the inevitable causal relationship between events and conditions, resulting in the discreteness of the results. For example, the load on the machine, the working state it is in, the performance of the parts and materials, the life of various parts, and the results of each test conducted under the same conditions are different.
Fuzziness is also a type of uncertainty that is much studied today. Fuzziness refers to an objective attribute in which the concept of the thing itself is unclear, there is no exact definition in quality, and there is no exact limit in quantity. For example, "good" and "bad", "excellent" and "inferior" in the evaluation of innovative design of mechanical product principle solutions, "normal" and "fault", "stable" and "unstable", " "Safe" and "unsafe", etc., all of the above show unclear boundaries, that is, there is a continuous transition process from one side of the difference to the other, presenting a "this is also that" state, that is, the members included It is difficult to determine whether it conforms to the concept of this thing, that is, it is vague which objects a set contains.
The uncertainty of information refers to the imperfect information caused by the limitations of objective conditions, which results in information that decision-makers cannot know for certain when making decisions. For example, due to measurement difficulties, accurate data on wear, gluing, fatigue pitting, etc. of a certain part during operation cannot be obtained. This is the imperfection of objective information. In addition, the imperfection of subjective knowledge is caused by the insufficient evidence held by the decision-maker, that is, the imperfection of subjective knowledge, which is not enough to determine the actual status and quantitative relationship of things, resulting in unclear subjective understanding. sex. For example, in the development and design of mechanical products, due to the priority given to technological development, the future performance of the product cannot be fully controlled.
picture
Characteristics of the above three types of uncertainties:
First of all, they all have two sides. On the one hand, this uncertainty is an objective manifestation of things, such as the physical properties of materials, mechanical properties of materials, dimensional tolerances of components, etc.; on the other hand, it reflects people's level of understanding things. Uncertainties in subjective judgments such as knowledge level and policy level, decision-making ability, experience, intelligence and creativity, etc., such as the approximation of loads, distribution assumptions of structural fatigue strength, the degree of reality of the model, working conditions, etc.;
Secondly, some things show a single uncertainty, that is, one of randomness, fuzziness, and uncertainty of information. However, some things show two or three types of uncertainty at the same time;
Finally, the duality and unity of certainty and uncertainty. In mechanical design, a large number of phenomena and factors can be described and processed using traditional precise mathematical methods, definite mathematical models, precise measurements, and definite values. This kind of deterministic quantification and accuracy has always been the key to mechanical design. the goal pursued.
Due to the development of science and technology, especially the development and penetration of information science, the trend of interdisciplinary interdisciplinary and comprehensive is increasingly strengthened. The requirements for the depth and breadth of the problems to be solved are getting higher and higher, and more complex technical problems are required to be solved. And to solve related economic, political, resource, environmental and other issues, there are many uncertainties in these fields, precise mathematical methods are powerless, and fuzzy mathematics, convex set models and other inexact Mathematical methods can well describe and deal with such non-deterministic problems.
