14-08-2012, 04:04 PM
A novel evolutionary algorithm for dynamic economic dispatch with energy saving and emission reduction in power system integrated wind power
A novel evolutionary algorithm.pdf (Size: 1.55 MB / Downloads: 40)
Introduction
Economic Dispatch (ED) problem is one of the most important
problems in power system operation. It involves meeting the load
demand atminimum total fuel cost while satisfying various unit and
system constraints. The ED model is a nonlinear optimization
problem that considers its nonlinear characteristics including
discontinuous Prohibited Zones, Power Balance constraints, Generation
Limit constraints, Valve Point Effects constraints, Ramp Rate
limits, Spinning Reverse and cost functions.
This paper studies the economic dispatch problem for power
systems that contain a wind farm. The economic dispatch for power
systems can be divided into traditional static economic dispatch
and Dynamic Economic Dispatch (DED). The static optimal dispatch
seeks to achieve an optimal objective for the power system at
a specific time, but will not take into account the intrinsic link
between the systems at different time moments. The dynamic
optimal dispatch takes into account of the coupling effect of system
at different time moments, such as the limit on the generator
climbing rate. As a result, its computation process is more complex
than that of a static optimal dispatch, but the computation results
are more in line with actual requirements.
The quantum chaotic genetic algorithm concept
Quantum Evolutionary Algorithm (QEA) is an evolutionary algorithm
based on the quantum computing concept. It incorporates the
concept in quantumcomputing and adopts the unique coding format
to achieve better experimental results on the combinatorial optimization
problems. However, when dealing with the high dimensional
multi-modal function optimizationproblem, it is prone to fall into the
local optimum and its computational efficiency is poor.
For the above-mentioned deficiencies of QEA, this research
integrated the global optimization ability of genetic algorithm, local
searching capability based quantum probability model, the sensitive
dependence of chaotic algorithm to initial value, and the traverse of
the search space to establish a new improved quantum evolution
algorithm, that is “Chaotic Quantum Genetic Algorithm” (CQGA).
Case study 3
The daily total power generation cost comparison charts
between the algorithm proposed in this paper and four other
algorithms (DP, GA, EP and QGA) for seven different days as shown
in Fig. 6 (data for the calculation are: 10 conventional thermal
power generators and 20 wind turbine generator systems of the
same model type; their parameters and computation conditions
are the same shown as in Case study 1). Comparing the daily
average of the total power generation cost within the seven days in
the charts, using the algorithm proposed by this paper can save
about 12e21% of the total power generation cost at Day 1 than
using the other four methods (the other six comparison results are:
at Day 2 is from 16% to 25%, at Day 3 is from 14% to 26%, at Day 4 is
from 15% to 27%, at Day 5 is from 13% to 25%, at Day 6 is from 12% to
23% ad at Day 7 is from 11% to 21%). Fig. 7 is the comparison chart
for the CPU computation time between the algorithm proposed in
this paper (CQGA) and four other algorithms. It can be shown from
the chart (the running condition is the same settle as shown in
Fig. 6 at Day 3) that using CQGA can save about 35%, 20%,16% and 9%
of the CPU computation time than using the four other methods
(that is DP, GA, EP and QGA method serially).