Industrial products inventory management using deterministic and probabilistic models.
Zapciu, Miron ; Zapciu, Georgeta ; Georgescu, Luminita 等
1. INTRODUCTION
Having the stocks in industrial production requires costs with
supply or production, costs of specific storage (storage, maintenance,
handling, conservation, etc.) and others losses due to depreciation in
special cases. To manage stocks are managed stock entries and exits or
supplies in stock. Supplies of products could be modelled using three
possible situations of exits from stock (Krajewski & Ritzman, 2001):
deterministic (calculated as the known functions); random with a known
statistical distribution; unknown variable of stock exit.
The main activities involved in inventory management are the
following: the size of lots of products, stock level, the volume of
replenishment orders, the storage costs, penalty costs, costs of the
order etc. For an optimization of inventory levels, the industrial
management requires most often, a criterion of minimum total cost of
these stocks. Items on which we have intervention are: optimal level of
stocks, the optimal volume of orders for replenishment, or the optimal
time for replenishment, frequency of replenishment, minimum total cost.
For optimum management of stocks are using different mathematical
models, shared depending on the nature of demand in deterministic and
probabilistic models.
2. DETERMINISTIC MODELS FOR INVENTORY MANAGEMENT
2.1 Storage model of constant demand, constant replenishment period
and lack of stock
This model considers the following assumptions: storing a single
type of product whose consumption is a linear function during time,
demand of the product is N pieces for a period [theta], replenishment of
stock is performed instantaneously (in theory) at equal time (T) and
equal quantity (q). Unit cost of storage in unit time is [c.sub.s], the
cost to launch a replenishment order is the parameter [c.sub.lan] (the
total cost of the replenishment operation that does not depend on q) and
is not allowed lack of stocks. Graphical representation of this model is
made in the Fig. 1.
Under the assumptions of this model, for each time period T are
performed following costs (equation 1):
C(q) = [c.sub.lan] + 0.5 * qT [c.sub.s] (1)
[FIGURE 1 OMITTED]
Considering the stock is equal with the average quantity of the
products q/2 and the equal periods T is calculating the number of
replenishment,y using the equation:
y = N / q = [theta]/T (2)
For this model objective function includes all costs related to
inventory management. Given the values of y in the equation (2) is
obtaining the following expression for the objective "total
cost" function:
C(q) = 1/q N [C.sub.lan] + 1/2 q c, [theta] (3)
The quantity [q.sub.opt] of optimal replenishment, the optimal
number of replenishment [y.sub.opt] and the optimal time between two
replenishments [T.sub.opt] are calculated using the equation (4). Since
C"(q)> 0 extreme value is a minimum (because q has only positive
values).
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)
2.2 Storage model using different types of products
This type of simple model is very close to the real situations from
industrial companies. Practical situations of most enterprises are those
in which stocking more products are made thanks to the diversification
need, resulting from the necessity to better meet customer requirements
(example: in automotive industry). For examples of this model is built
the stock model for the case using different products.
Assumptions are: k of Pi products are stored; [P.sub.i] = 1, 2,
..., k whose consumption is given by linear functions of time and the
request of [P.sub.i] is [N.sub.i]; for a period of time [theta],
replenishment is performed instantaneously at the intervals [T.sub.i],
in equal quantity of [q.sub.i]; unit storage costs are [c.sub.si]; the
costs of delivery [c.sub.lan,i], with i = 1,2 k.
In this case is not admissible for lack of any stock of the
[P.sub.i] products. According to equation (3), the total cost of storage
for the product [P.sub.i] is calculated using equation (5):
[C.sub.i]([q.sub.i]) = 1/[q.sub.i][N.sub.i][c.sub.lan,i] +
1/2[q.sub.i][theta][c.sub.si] (5)
[FIGURE 2 OMITTED]
Graphical representation for this type of model, where consumption
of products varies over time and the level of replenishment is less than
the maximum allowed, is shown in the Figure 2.
It should be noted that with the changing profile of manufacturing
in volume production or methods of organization and management of
production and supply is very important to reconsider and to study in
depth the management of stocks (Zapciu, 2006).
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)
The equation (6) allows obtaining the following important
parameters: [q.sub.opt]--optimal batch; [y.sub.opt]--optimal frequency
of orders, [C.sub.opt]--optimum cost.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)
Optimal management is reprezented by the minimum cost. This
parameter is determined by the equation (7).
3. PROBABILISTIC MODEL FOR THE COSTS OF THE INVENTORY MANAGEMENT
Using a probabilistic model for the stocks, was analyzed two
significant cases: first case is the stock model for a product with
stochastic market demand, with neglected cost to depose the product,
with supplementary cost when it is no stock available; the second case
considers a product with stochastic market demand, with knew deposed
cost and penalty cost when the product is not always available.
The second stock cost model studied was more complex and it is to
close to the most real industrial cases. When the demand of the product
is a continuous function f(x), the total cost of the stocks C(,), is
calculated using the relations:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)
where:
[c.sub.s] is unitary cost of stock; [c.sub.p]--unitary penalty cost
for unavailability of the product; s--instantaneously level of the
stock.
C(s) function get the minimum in the point [s.sub.opt] when stock
unavailability L(s) = [rho], according to the equations:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (9)
To exemplify the application of this model to stock administration
is presented a case study calculus; the model considers a stocked
product with stochastically and continuously demand, with storage cost
and penalty cost when the product is not available.
Considering the example when the function for the demand of the
product is [F.sub.x]) = (-1/2 x +3), considering the realization of 10
products batch size, unitary cost to stock the product is 1.5 [micro]m,
unitary penalty cost when the product is not available is equal with 0.5
[micro]m, consequently, optimal value for the stock cost is 162.9
[micro]m.
In the case when the function of the demand of the product has a
sinus component, for example g(x) = (-0.5x+3) +sinx, the total stock
cost in this case is f(x) = 141.1 [micro]m.
4. CONCLUSION
Inventory management is one of the issues particularly important if
the production requires a lot of products to be in stock. New result of
this paper is mathematical models presented that can be used in real
cases of industrial production. For example, applying the models in the
case study from an SME's roumanian company, considering the data
entry [c.sub.s] = 1.5 and [c.sub.p] = 0.5, the total stock costs of 20
batch products where: C(s) = 154.4 for the model with random discrete
demand and C(s) = 162.9 for the model with random continues demand.
Using a flexible manufacturing system, the product flow could be
variable and more adapted to the market product demand. Practically, the
stock level is calculated taking into account that the total cost for a
product series to be minimum.
Mathematical model presented for the cost of the stock could be
used for any other industrial case, but with the remark that must be
knew the time variation of the function for the market demand of the
product.
Next step will be finding new hybrid stock management models more
adapt to the industrial company and helping to become competitive into a
difficult market.
5. REFERENCES
Ispas, C.; Zapciu, M. & Mohora C. (2003). Quality ofthe product
in the context of flexible manufacturing, Proceedings of the
International Symposium on Quality and Standardisation, pp. 236-242,
Gheorghe G (Ed.) ISBN 9738177-42-1, Bucharest, October 2003, Bren
Edition, Bucharest
Krajewski, L.; Ritzman, L. (2001). Operations Management: Strategy
and Analysis, 2nd edition, Addison-Wesley, ISBN 1-890367-01-X,
Manchester
Zapciu, G.V. (2006). Contributions about the application of project
management concepts in material flow optimisation in industrial
companies. PhD Thesis, University POLITEHNICA of Bucharest
Zapciu, G.V. & Zapciu, M. (2005). Taking into account the risk
in logistics problems, Proceedings of International Conference
Energy-Environment--CIEM, pp. 423-428, ISBN 975-86948-5-X., Bucharest
November 2005, Bren Edition, Bucharest
Zapciu, M.; Cote^, E.C. & Stefan, S. (2001). Techniques of
management, Bren Edition, ISBN 973-8143-21-7, Bucharest
*** (2009) http://www.stokecoll.ac.uk/ Paper Webpage e032.htm,
Accessed on:2009-04-22