you summarise 100–500 pages into a one-page report with the right
recommendation?
The selection of the optimal trade-offs between multiple public-policy objectives
must, by definition, require the participation of responsible political decision makers.
They are required to specify these optimal trade-offs among objectives in a political
environment that is dynamic and changing and very demanding of their time.
14.6 Formulation of Planning Alternatives
A variety of modelling techniques are appropriate to the generation of efficient
infrastructure plans. In general, the approaches may be categorized as optimization
or simulation models. Optimization models can be applied to problems in which the
technical feasibility restrictions g i (X) ¼ 0 and the objective functions F j (X) are both
explicit and mathematically well defined. Constraints which can be approximated by
linear functions are particularly desirable since they facilitate the use of linear and
separable programming methods.
Simulation methods are used extensively in infrastructure planning. In its most
basic form, simulation modelling consists of finding solutions to the feasibility
conditions g i (X) = 0, i ¼ 1, 2, . . ., m, and evaluating objectives F j (X) = 0, j ¼ 1,
2, . . ., n, in either a deterministic or probabilistic form. The varied nature of
simulation models defies simple categorization. Since simulation methods will not,
in general, produce Pareto efficient plans, optimization and simulation are sometimes used conjunctively. The planning problem is first described by simplified
models which are amenable to optimization. The generated plans are then studied
using more realistic (and complex) simulation models. The optimization stage is
considered as a screening method for identifying promising (Pareto efficient) alternatives, and the simulation stage provides a realistic evaluation of plan feasibility
and performance.
Optimization models generally require a number of simplifying assumptions, and
as such, are seldom exact models of real-world planning situations. As opposed to
simulation methods, however, optimization can be used to approximate the feasibility frontier or objective transformation surface. Of course, the accuracy of the plans
generated will depend on the modelling assumptions, and as previously mentioned,
different optimization models of the same problem may not produce identical
transformation surfaces.
Perhaps the most common optimization model is the modified target planning
model discussed in the previous section. A feasibility frontier is generated by
maximizing a single objective F 1 (X) while varying the target levels T j of other
objectives F j (X):
Maximize F 1 X
ð Þ
ð14:5Þ
14.6 Formulation of Planning Alternatives
395
recommendation?
The selection of the optimal trade-offs between multiple public-policy objectives
must, by definition, require the participation of responsible political decision makers.
They are required to specify these optimal trade-offs among objectives in a political
environment that is dynamic and changing and very demanding of their time.
14.6 Formulation of Planning Alternatives
A variety of modelling techniques are appropriate to the generation of efficient
infrastructure plans. In general, the approaches may be categorized as optimization
or simulation models. Optimization models can be applied to problems in which the
technical feasibility restrictions g i (X) ¼ 0 and the objective functions F j (X) are both
explicit and mathematically well defined. Constraints which can be approximated by
linear functions are particularly desirable since they facilitate the use of linear and
separable programming methods.
Simulation methods are used extensively in infrastructure planning. In its most
basic form, simulation modelling consists of finding solutions to the feasibility
conditions g i (X) = 0, i ¼ 1, 2, . . ., m, and evaluating objectives F j (X) = 0, j ¼ 1,
2, . . ., n, in either a deterministic or probabilistic form. The varied nature of
simulation models defies simple categorization. Since simulation methods will not,
in general, produce Pareto efficient plans, optimization and simulation are sometimes used conjunctively. The planning problem is first described by simplified
models which are amenable to optimization. The generated plans are then studied
using more realistic (and complex) simulation models. The optimization stage is
considered as a screening method for identifying promising (Pareto efficient) alternatives, and the simulation stage provides a realistic evaluation of plan feasibility
and performance.
Optimization models generally require a number of simplifying assumptions, and
as such, are seldom exact models of real-world planning situations. As opposed to
simulation methods, however, optimization can be used to approximate the feasibility frontier or objective transformation surface. Of course, the accuracy of the plans
generated will depend on the modelling assumptions, and as previously mentioned,
different optimization models of the same problem may not produce identical
transformation surfaces.
Perhaps the most common optimization model is the modified target planning
model discussed in the previous section. A feasibility frontier is generated by
maximizing a single objective F 1 (X) while varying the target levels T j of other
objectives F j (X):
Maximize F 1 X
ð Þ
ð14:5Þ
14.6 Formulation of Planning Alternatives
395