Wartime influence factor and epsilon-SVR-based battle damage spare part prediction method

Abstract

The invention relates to a wartime influence factor and epsilon-SVR-based battle damage spare part prediction method. The method comprises the following steps of: designing an overall process on the basis of a wartime influence factor analysis and epsilon-SVR-based battle damage spare part demand prediction method; carrying out UML-based wartime influence factor analysis, ER map-based wartime influence factor concept modeling and social network analysis-based important factor extraction and important factor quantification; constructing a wartime influence factor-spare part demand-based epsilon-SVR prediction; and carrying out prediction by using the epsilon-SVR model. Aiming at the problem that the present common spare part demand prediction methods cannot satisfy wartime equipment damagenonlinearity and law unknowns and are small in data size, the wartime influence factor analysis and epsilon-SVR-based battle damage spare part demand prediction method is high in feasibility and effectiveness, and is capable of providing scientific decision basis for equipment support work.

Description

technical field [0001] The invention relates to a demand forecasting method, in particular to a wartime spare parts forecasting method based on wartime influencing factors and ε-SVR, and belongs to the technical field of computer equipment support simulation technology. Background technique [0002] Computer equipment support simulation has become an important method and means to research and solve equipment support problems. With the transformation of the form of war from mechanization to informatization, the complexity of equipment battle damage has further increased. It is of great significance to scientifically and accurately predict the type and quantity of wartime equipment maintenance equipment. [0003] UML (Unified Modeling Language) is a language for visualizing, specifying, constructing and documenting artifacts in software-intensive systems. UML was initiated by world-renowned object-oriented technical experts G.Booch, J.Rumbaugh and I.Jacobson. Based on the Boo...

AI Technical Summary

Problems solved by technology

On the one hand, traditional prediction methods require a large number of samples, or the prior probability of equipment damage, or the loss of spare parts has a stable trend; on the other hand, due to the complexity, randomness, nonlinearity and uncertainty of wartime equipment damage In addition, the damage probability or consumption pattern of equipment in wartime is different from that in normal times, and the data of equipment damage in battle is also difficult to obtain, which leads to certain limitations in the use of these traditional methods.

Images