L1 2 3 Modeling LP

ME 421: Industrial
Engineering and Operations

Research
Lecture Plan
Lecture Content Lecture Content

Lecture 1 Introduction to OR, LP Lecture 21 Queuing theory
Lecture 2 LP Lecture 22 Queuing theory
Lecture 3 Graphical Method Lecture 23 Production Planning and Control
Lecture 4 Simplex Method Lecture 24 Production Planning and Control
Lecture 5 Simplex Method Lecture 25 Aggregate production planning
Lecture 6 Simplex Method Lecture 26 Product design, Value analysis and value engineering
Lecture 7 Simplex Method Lecture 27 Plant location and layout
Lecture 8 Dual Problem Lecture 28 Job, batch, and flow production methods
Lecture 9 Dual Simplex Method Lecture 29 Group technology
Lecture 10 Dual Simplex Method Lecture 30 Inventory control, EOQ, EPQ
Lecture 11 Dual Simplex Method Lecture 31 Inventory control, EOQ, EPQ
Lecture 12 Concept of unit worth of resource Lecture 32 Inventory control, EOQ, EPQ
Lecture 13 sensitivity analysis Lecture 33 Inventory control, EOQ, EPQ
Lecture 14 Transportation problems Lecture 34 Forecasting
Lecture 17 Assignment problems Lecture 37 Scheduling and loading
Lecture 18 Assignment problems Lecture 38 Scheduling and loading
Lecture 19 Network models: CPM and PERT Lecture 39 Scheduling and loading
Lecture 20 Network models: CPM and PERT Lecture 40 Line balancing
Lecture 41 Break-even analysis
Lecture 42 Supply chain management
Lecture 43 Industry 4.0
Course Resources
• Texts:
1. S. L. Narasimhan, D. W. McLeavey, and P. J. Billington, Production,
Planning and Inventory Control, Prentice Hall, 1997.
2. J. L. Riggs, Production Systems: Planning, Analysis and Control, 3rd Ed.,
Wiley, 1981.
• References:
1. A. Muhlemann, J. Oakland and K. Lockyer, Productions and Operations
Management, Macmillan, 1992.
2. H. A. Taha, Operations Research - An Introduction, Prentice Hall of India,
1997.
3. J. K. Sharma, Operations Research, Macmillan, 1997.
Lectures 1, 2 & 3:
Introduction, Modeling to LP
1. Operations Research Models
Operations Research Models
Operations Research Models
• Wire of L inches.
• Make a rectangle from the wire such that area of rectangle should
be maximum.
• Alternates
• Width = w
• Height = h
• Restriction
• 2(w+h) = L
• w,h ≥ 0
• Objective function: Maximize z = wh
Search and Optimization
A task of searching for a set of decision variables which would minimize or
maximize objective function(s) subject to satisfying constraints and bounds on
decision variables.
• Decision variables: (x,y)

• Objective function: f(x,y)
• Constraints: g(x,y) or h(x,y)
• Optimization Modeling
Minimize f(x,y)
g(x,y) ≥ 0
h(x,y) = 0
xl ≤ x ≤ xu
yl ≤ y ≤ yu
Solutions
• Feasible solution
• A solution satisfies all the constraints.
• Optimal solution
• A feasible solution which yields the best value of objective function in the
entire feasible search space
• Sub-optimal or local optimal solution
• The feasible solution which is optimal in its vicinity.
• Search space
• The space defined by the constraints and limits on the variables.
• Feasible search space
• The space in which any point in it is always feasible.
Steps in an Optimization Task
• Need for optimization
• Problem formulation or modeling
• Identify problem parameters
• Choose design variables from parameters
• Formulate constraints
• Formulate objective function
• Set up variable bounds
• Requires 50-60% of the effort
• Choose an optimization algorithm
• Obtain solution
• Reformulation and rerun, if desired
Design Variables and their Bounds
• List any and every parameter related to the problem
• Identify parameters sensitive for the given design or problem
• Specify the type of each parameter (binary, discrete, real)
• Choose few of them as design variables
• First thumb rule: Use as few variables as possible
• Usually from the experience of the user
• From minimum variability consideration
• From sensitivity analysis etc.
• Bounds or limit on decision variables
xl ≤ x ≤ xu
yl ≤ y ≤ yu
Constraints
• Represent limit on resource or on certain physical phenomenon, for
example satisfy stress and deflection limitation
• Inequality constraint: g(x) ≥ 0 or g(x) ≤ 0
• mostly encounter in engineering design problems
• Equality constraint: h(x) = 0,
• In linear programming (LP), or satisfying demand etc.
Objective Function
• Minimize or maximize
• Optimization methods are generally developed for minimization
• Use duality principle
• Min f(x) = - Max f(x)
Solving OR Models
• Linear Programming
• Integer Programming
• Dynamic Programming
• Network Programming
• Nonlinear Programming
• OR Techniques do not find solutions in closed form
• Algorithm
• Fixed computational rules that are applied repetitively to the problem
• Repetition is called iteration
• Each iteration a solution is getting closer to the optimum.
• Tedious and Voluminous
• Executed on the computer
Thank you.
2. Modeling with Linear Programming (LP)
• Two-variable LP Model
Alternatives or
Decision variables?
Constraints
Objective function
Modeling: Reddy Mikks Problem
• Decision variables
• Objective function
• Constraints
• Subject to
• Feasible solution: A solution (x1, x2) satisfying all constraints

• Infeasible solution: Otherwise
• Check
• Solution (3,1)
• Another solution (4,1)
Properties of LP Model
• Objective function and constraints are linear. Linearity implies that
LP must satisfy three basic properties
• Proportionality
• Contribution of each decision variable in both the objective function and constraints
is directly proportional to the value of variable.
• Additivity
• Total contribution of all decision variables in the objective function and in the
constraints to be the direct sum of the individual contribution of each variable.
• Certainty
• Coefficients of the objective function and the constraints are deterministic.
Graphical Solution to Reddy Mikks Model
• Determination of the feasible solution space
• Determination of the optimum solution among all the feasible
points in the solution space.
Reddy Mikks Model: Determine the feasible space
• Determine optimal solution
E D
C
A
B
• Corner Solution
• Find the optimum solution from the corner solutions only.
Diet Model
• Decision Variables
Diet Model
• Objective function:
• Constraints
• Variable bound
Graphical Solution to Diet Model
LP Applications
• Many examples are given in the book by TAHA (Self reading)
• Selected LP applications
• Production Planning and Inventory Control
• Bus Scheduling
• Trim Loss
Production Planning and Inventory Control Model
• Decision Variables
• Objective function
Demand
Nonnegative condition
• Constraints
Bus Scheduling Model
Bus Scheduling Model
Trim Loss or Stock Slitting Model
• Setting knife to the desired width

• A case of 3 settings as shown in Fig.
• Example of feasible combination
• Slit 300 rolls using setting 1 and 75 rolls using setting 2
• Slit 200 rolls using setting 1 and 100 rolls using setting 3
• Which combination is better?
Trim Loss or Stock Slitting Model
• Trim loss as evaluation of goodness of solution
• Assuming the standard roll is of length L
• Trim loss
• Combination 1: 300 (4XL) + 75 (3XL) = 1425 ft2
• Combination 2: 200 (4XL) + 100 (1XL) = 900 ft2
• Surplus of 5-, 7-, and 9-ft rolls must be considered in the computation of
trim loss
• Combination 1: Surplus of 300-200 = 100 extra 7-ft rolls in setting 1 and 75 extra
7–ft rolls in setting 2. Total waste = 175 (7XL) = 1225 ft2.
• Combination 2: Setting 3 produces 200-150 = 50 extra 5-ft rolls and total waste
is 50 (5XL) = 250L ft2.
• Total trim loss area is
• Combination 1: 1425L + 1225L = 2650L ft2.
• Combination 2: 900L + 250L = 1150L ft2. (BETTER)
Mathematica Model: Trim Loss
• Minimize trim-loss area by satisfying the demand.
• Promising setting cannot yield a trim-loss roll of width 5 feet or
larger.
Mathematica Model: Trim Loss

L1 2 3 Modeling LP

Uploaded by

Copyright:

Available Formats

L1 2 3 Modeling LP

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

L1 2 3 Modeling LP

Uploaded by

Copyright:

Available Formats

ME 421: Industrial

Engineering and Operations

Lecture Content Lecture Content

• Decision variables: (x,y)

• Feasible solution: A solution (x1, x2) satisfying all constraints

• Setting knife to the desired width

You might also like