Sei sulla pagina 1di 3

Table 1.

1: Comparison table of the optimization models between similarities and


differences of LP, NLP and QP.

Optimization Linear Non-linear Quadratic Programming,


models Programming, LP Programming, NLP QP
Objective Linear objective Non-linear objective Non-linear objective
Function function function function
Constraints Linear constraints Linear or non-linear Linear constraints
constraints
Gradient Convex Convex and non-convex Convex and non-convex
preformed
Vector of Decision variables Decision variables Decision variables raised
decision only raised to a raised to a power or to a power or multiplied
variable power with 1. multiplied by other by other variables.
variables or logarithms
Advantages - Most suitable to - Applicable in linearity - Applicable in linearity
solve complex or non-linearity of or non-linearity of
problems relations of decision relations of decision
- Ease of use or variables variables
solve - Lower efficiency - Likely with the LP due
- Efficiency compare to LP to linear constraints
- Fast and reliable to - Lower efficiency
solve compare to LP

Disadvantages - Only applicable in - Sometimes will be - Sometimes will be


linearity of relations performed in non- performed in non-convex
of variable convex condition condition

Table 1.2: Comparison table of the optimization models between similarities and
differences of direct search and stochastic search.

Optimization Direct Search Stochastic Search


models
Objective Non-linear Random objective
Function objective function function
Constraints Non-linear or linear Non-linear or linear
constraints constraints
Gradient - No require any Convex and non-convex
preformed information of
gradient
Advantages - Suitable to solve - Suitable to solve non-
non-differentiable differentiable objective
objective function function or non-even
or non-even continuous
continuous - Random iteration to be
- To solve difficult used
global optimization
problems
- Ease to solve, high
efficiency

Disadvantages - Need to do few - Uneasy to solve


times of iteration - Not always can find the
global optimum point
- Needed higher
computational effort

Table 1.3: Comparison table of the optimization models between similarities and
differences of LP, NLP and QP.

Optimization Integer Mixed integer linear Mixed integer non-linear


models programming, LP programming, MILP programming, MINLP
Objective Linear objective Linear objective function Non-linear objective
Function function function
Constraints Linear constraints Linear or non-linear Linear or non-linear
constraints constraints
Gradient Convex Convex Convex and non-convex
preformed
Vector of Decision variables Decision variables raised Decision variables raised
decision only raised to a to a real integer and to a real integer and
variable integer value numerical mathematical numerical mathematical
Advantages - Suitable to solve - More easier to solve - Applicable in linearity
discrete problem compare to MINP or non-linearity of
- Applicable for relations of decision
linear objective variables
function and
constraints
Disadvantages - Only applicable in - Only applicable in - Sometimes will be
linearity of relations linearity of relations of performed in non-convex
of variable variable condition
- The vector of - The vector of decision - The vector of decision
decision variables variables include an variables include an
must be an integers integers value integers value

Potrebbero piacerti anche