Stochastický model logistického systému výrobní firmy se zaměřením na řízení zásob

Abstract

The issue of inventory management is very relevant nowadays because this important managerial activity can affect significantly other business processes, flexibility and competitiveness of company. Considerable turbulence of the market environment and impact of many stochastic phenomena may complicate this activity and affect negatively.

One of the basic methodological approaches to inventory management is an approach of optimizing. The basic problem is usually the optimizing of cost type – consequently the minimizing of the total inventory costs. To solve optimization tasks, we can use either mathematical and statistical tools of inventory theory, or simulation which is an effective and flexible tool of quantitative methods to support of managerial decision-making.

The issue of inventory management is solved not only as an isolated process at the submitted dissertation but as an integral part of the conceptual model of logistic system of ficticious manufacturing company. Its creation is a benefit and one of the main objectives of the presented dissertation.

The logistic system is researched and analyzed on two levels:  the level of partial logistic processes (e.g. subsystem of customer order management, production planning subsystem, material supply subsystem, inventory management subsystem, distribution subsystem etc.) which is researched in particular by a heuristic stochastic simulation Monte Carlo;  the level of global logistic system which is researched by the known method – Cost and Revenue Analysis. This quantitative analysis of the company effectiveness is based on knowledge which are consistent with conclusions of the current economic theory. In view of the complexity of researched processes there have not been used classical mathematical optimization models at the context of dissertation. One of the other benefits of the submitted dissertation is the stochastic simulation approach to optimization on both of above-mentioned levels.

Analyses of these processes are based on the commonly used statistical methods and procedures. For computer modeling of the partial processes of logistic system was used widely available software MS Excel. Although this software is limited for modeling the complex real systems, it is completely sufficient for the purposes of stochastic approach to solve optimization problems.

The results of simulation were used for creating of a profit matrix which represents the final part of simulation modeling. This matrix is expressed in three versions:  matrix of mean values of profits/losses;  matrix of corresponding standard deviations of profits/losses;  matrix of minimum expected profits/losses with predetermined probability (Decision Matrix).

The importance of this matrix is mainly in the fact that it allows to find out a certain manoeuvrable area and ways for improving efficiency – e.g. determination the point of global optima (global maximum of profit), finding points of limited profit optimum at the case of technical-economic capacity constraints, take a decision about the possible position and value of profit when company wants to get into the new markets etc. At decision analysis of various risk variants is taken into account the characteristics of manager personality and in particular his attitude to risk.