Introducción
[Link]. Edgar O. Reséndiz-Flores
TECNOLÓGICO NACIONAL DE MÉXICO
Instituto Tecnológico de Saltillo
[Link]. Edgar O. Reséndiz-Flores Introducción
Información del curso
[Link]. Edgar O. Reséndiz-Flores
e-mail: [Link]@[Link]
[Link]@[Link]
Evaluación:
Dos examenes: 30 %
Tareas: 50 %
Exposiciones: 20 %
[Link]. Edgar O. Reséndiz-Flores Introducción
Información del curso
Contenido general:
Introducción: algebra lineal, análisis convexo
Programación lineal: método simplex, de la gran M, de dos fases y
punto interior
Teorı́a de la dualidad y sensibilidad: problema dual, algorı́tmo dual
simplex
Programación entera: modelación matemática usando variables enteras
y binarias, método branch and bound, método de corte de planos
Programación por metas: modelación matemática de problemas
multi-objetivo, programación por objetivos no secuencial y secuencial
Transporte y asignación: definición del problema de transporte y
asignación, soluciones básicas, método simplex del transporte, análisis de
sensibilidad, método Húngaro.
Programación no lineal: modelación de problemas de inventarios,
funciones concavas y convexas, problemas sin restricciones y el método
del gradiente, programación cuadratica, técnica de maximización
secuencial irrestricta.
[Link]. Edgar O. Reséndiz-Flores Introducción
Operations Research
Definition
[Link]. Edgar O. Reséndiz-Flores Introducción
History of Operations Research
1881: F. W. Taylor, performed his time studies, What is the best way to
do a job?
1909: Henry L. Grant, bar charts for scheduling problems and Agner
Krarup Erlang introduced the discipline of queuing for Telephone
exchange in Copenhagen,
1913: F. W. Harris developed the economic order of quantity for
inventory management,
1920: John Von Neumann introduced the theory of games,
1930: Leontif’s input-output models and Kantarovich’s mathtematical
programming models for the Soviet economy,
1940: Hitchcock’s transportation problem, Stigler’s diet planning and
advances based on military applications,
1947: George Bernard Dantzig developed the simplex method for linear
programming,
1950: John F. Nash extended von Neumann’s results in game theory and
proved some main theorems, Bellman’s dynamic programming principle
1951: Khun and Tucker’s optimality conditions for nonlinear optimization
problems (reinvention of work by Karush in 1939). First computer-based
simplex method.
[Link]. Edgar O. Reséndiz-Flores Introducción
Proceso de modelación matemática
1 Problem recognition ”We have done it always like that” and
the sister in crime ”We have never done it like that”
2 Authorization to model
3 Model building and data collection
4 Model solution
5 Model validation
6 Model presentation
7 Implementation
8 Monitoring and control
[Link]. Edgar O. Reséndiz-Flores Introducción
Some applications
[Link]. Edgar O. Reséndiz-Flores Introducción
Production planning: problem description
Suppose there are three products to be manufactured P1 , P2 and P3 .
The three products sell for 20, 15 and 17, respectively.
They are manufactured in two machines M1 and M2 .
Each of these products has to be processed on both machines where the
order is irrelevant.
M1 has capacity of 8 hours. M2 has capacity of 9 hours. After that time,
they require regular maintenance. Machines shut off and are no longer
available.
Processing times are as follows
Suppose there is a clock attached to each machine and it shows the
processing time left before every maintenance.
[Link]. Edgar O. Reséndiz-Flores Introducción
Production planning: problem description
Production costs are 120 per hour of machine time on M1 , and 90 per
hour of machine time on M2 . That is, 2 and 1.50 per minute, respectively.
The processing costs of the three products on M1 are 6, 10 and 8,
respectively. While on M2 are 9, 1.50 and 4.50.
The overall unit processing cost are 15, 11.50 and 12.50 for P1 , P2 and
P3 .
Considering the selling prices, we have profit contributions of 5, 3.50 and
4.50 per unit of each of the three products.
[Link]. Edgar O. Reséndiz-Flores Introducción
Production planning: mathematical modeling
Objective function:
Defining x1 , x2 and x3 the amount of P1 , P2 and P3 that will be
manufactured and sell.
The objective function is to maximize the profit, which is the sum of
profits for each product, that is,
max z = 5x1 + 3.5x2 + 4.5x3
Constraints:
We need to formulate one constraint for each machine.
It should state that the actual usage of the processing time does not
exceed the available one which is known in advance.
For M1 : 3x1 minutes + 5x2 minutes + 4x3 minutes should not exceed
540 minutes. Thus,
3x1 + 5x2 + 4x3 ≤ 540
For M2 : 6x1 minutes + 1x2 minutes + 3x3 minutes should not exceed
480 minutes. Thus,
6x1 + 1x2 + 3x3 ≤ 480
[Link]. Edgar O. Reséndiz-Flores Introducción
Production planning: mathematical modeling
Model:
Solution:
[Link]. Edgar O. Reséndiz-Flores Introducción
Employee Scheduling: problem description
[Link]. Edgar O. Reséndiz-Flores Introducción
Employee Scheduling: mathematical modeling
Model:
Solution:
[Link]. Edgar O. Reséndiz-Flores Introducción
Dynamic production: problem description
[Link]. Edgar O. Reséndiz-Flores Introducción
Dynamic production: mathematical modeling
Objective function & Constraints:
[Link]. Edgar O. Reséndiz-Flores Introducción
Dynamic production: mathematical modeling
Model:
Solution:
[Link]. Edgar O. Reséndiz-Flores Introducción