Abhishek Kumar
ABSTRACT
Most of the real-world black-box optimization problems are associated with multiple non-linear as well as non-convex constraints, making them difficult to solve. In this work, we introduce a variant of the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) with linear timing complexity to adopt the constraints of Constrained Optimization Problems (COPs). CMA-ES is already well-known as a powerful algorithm for solving continuous, non-convex, and black-box optimization problems by fitting a second-order model to the underlying objective function (similar in spirit, to the Hessian approximation used by Quasi-Newton methods in mathematical programming). The proposed algorithm utilizes an e-constraint-based ranking and a repair method to handle the violation of the constraints. The experimental results on a group of real-world optimization problems show that the performance of the proposed algorithm is better than several other state-of-the-art algorithms in terms of constraint handling and robustness.
Supplemental Material
Available for Download
Supplemental material.
- Nikolaus Hansen, Sibylle D Müller, and Petros Koumoutsakos. 2003. Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (CMA-ES). Evolutionary computation 11, 1 (2003), 1--18.Google Scholar
- Michael Hellwig and Hans-Georg Beyer. 2018. A matrix adaptation evolution strategy for constrained real-parameter optimization. In 2018 IEEE Congress on Evolutionary Computation (CEC). IEEE, 1--8.Google ScholarCross Ref
- Abhishek Kumar, Guohua Wu, Mostafa Z. Ali, Rammohan Mallipeddi, Ponnuthurai Nagaratnam Suganthan, and Swagatam Das. 2020. A Test-suite of Non-Convex Constrained Optimization Problems from the Real-World and Some Baseline Results. Swarm and Evolutionary Computation, (2020). Google ScholarCross Ref
Index Terms
- A modified covariance matrix adaptation evolution strategy for real-world constrained optimization problems
Recommendations
A modified Covariance Matrix Adaptation Evolution Strategy with adaptive penalty function and restart for constrained optimization
In the last decades, a number of novel meta-heuristics and hybrid algorithms have been proposed to solve a great variety of optimization problems. Among these, constrained optimization problems are considered of particular interest in applications from ...
A Multiobjective Particle Swarm Optimizer for Constrained Optimization
Constraint handling techniques are mainly designed for evolutionary algorithms to solve constrained multiobjective optimization problems CMOPs. Most multiojective particle swarm optimization MOPSO designs adopt these existing constraint handling ...
Solving constrained optimization problems by using covariance matrix adaptation evolutionary strategy with constraint handling methods
ICIAI '18: Proceedings of the 2nd International Conference on Innovation in Artificial IntelligenceNumerous constraint handling techniques were proposed in the past to be used with evolutionary algorithms (EA). According to the no free lunch theorem, there is no single algorithm that can consistently outperform over all other algorithms for all types ...
Comments