Soft Comput
DOI 10.1007/s00500-015-1606-8
METHODOLOGIES AND APPLICATION
A model for resource-constrained project scheduling using
adaptive PSO
Neetesh Kumar · Deo Prakash Vidyarthi
© Springer-Verlag Berlin Heidelberg 2015
Abstract Resource-constrained project scheduling prob-
lem (RCPSP) is an important, but computationally hard prob-
lem. Particle swarm optimization (PSO) is a well-known and
highly used meta-heuristics to solve such problems. In this
work, a simple, effective and improved version of PSO i.e.
adaptive-PSO (A-PSO) is proposed to solve the RCPSP. Con-
ventional canonical PSO is improved at two points; during
the particle’s position and velocity updation, due to depen-
dent activities in RCPSP, a high possibility arises for the
particle to become invalid. To overcome this, an important
operator named valid particle generator (VPG) is proposed
and embedded into the PSO which converts an invalid par-
ticle into a valid particle effectively with the knowledge of
the in-degree and out-degree of the activities depicted by the
directed acyclic graph. Second, inertia weight (ω) that plays
a significant role in the quick convergence of the PSO is
adaptively tuned by considering the effects of fitness value,
previous value of ω and iteration counter. Performance of
the model is evaluated on the standard benchmark data of
the RCPSP problem. Results show the effectiveness of the
proposed model in comparison to other existing state of the
art model that uses heuristics/meta-heuristics. The proposed
model has the potential to be applied to other similar prob-
lems.
Communicated by V. Loia.
N. Kumar · D. P. Vidyarthi (B )
School of Computer and Systems Sciences, Jawaharlal Nehru
University, New Delhi 110067, India
e-mail: dpv@mail.jnu.ac.in
N. Kumar
e-mail: dgoldneetesh15@gmail.com
Keywords Particle swarm optimization · NP-hard ·
Heuristics · Resource-constrained project scheduling
problem · Makespan
1 Introduction
Resource-constrained project scheduling problem (RCPSP)
is an important, but NP-hard problem in project planning
research Blazewicz et al. (1983). Though, in the past, a num-
ber of methods using linear programming, heuristics and
meta-heuristics have been proposed to solve the RCPSP,
these methods do not often produce good results and take
substantial amount of time to converge. In recent times,
researchers have intensified their interest towards optimal
solution for RCPSP using evolved heuristics and meta-
heuristics (Chen and Huang 2007; Valls et al. 2005) making
this an active research area.
RCPSP contains a set of activities with deterministic exe-
cution time, precedence relation among activities, accumu-
lative resources availability constraints and its consumption
by the activities. The objectives in RCPSP is to find a feasible
schedule with some quality characteristics such as optimal
makespan and computation time, response time and through-
put, such that all the constraints are satisfied.
In literature, many exact scheduling algorithms to solve
the RCPSP problem that uses linear programming approach
(Pritsker et al. 1969; Kaplan 1996; Klein 2006; Mingozzi
et al. 1998; Kone et al. 2011) are available. Out of these, the
algorithms proposed by Brucker et al. (1998) and Mingozzi
et al. (1998) seem to be the most effective and comprehen-
sive. The drawback of their algorithms is that it solves the
problem of small instances only (60 activities) in a satisfac-
tory manner. Their method succumbs to higher convergence
time for large instances as the solution search space increases
123