Statistical methods of making decision on repair of technical means

摘要

The problem of making decision on repair of technical equipment in the condition of uncertainty on the basis of statistical data is considered. Operating hardware requires timely repair. When equipment is stopped for repair, the company will lose profit. The entrepreneur has to decide to stop equipment for repairs without loss of earnings. It is proposed to use a classic criteria for decision-making under uncertainty. They are the Laplace, Wald, Savage or Hurwitz criteria. The application of these criteria is considered as an example of vehicles which may be in good, fair or poor condition. The use of classical criteria allows minimizing losses. In addition, an experienced entrepreneur has accumulated some statistics to assess condition of the vehicles. This statistic appears in the form of a matrix of transition probabilities. Depending on the condition of the transport, the entrepreneur can receive a monthly income or incur losses. It is represented as a matrix of income. The matrices of transition probabilities and income can be used to predict future monthly income. In this case, we have a multi-stage task of decision-making problem of the need to stop the vehicle for repair or continue operation. This approach to decisions making on the repair work can be applied for a variety of technical means.