Unit commitment using the ant colony search algorithm

Abstract

T h e paper presents a n A n t Colony Search Algorithm (ACSA)-based approach t o solve t h e unit commitment (UC) problem. This ACSA algorithm is a relatively new metaheuristic for solving h a r d combinatorial optimization problems. It is a population-based approach t h a t uses exploitation of positive feedback, distributed computation as well as constructive greedy heuristic. Positive feedback is for fast discovery of good solutions, distributed computation avoids early convergence, a n d t h e greedy heuristic helps find adequate solutions in t h e early stages of t h e search process. T h e ACSA was inspired from natural behavior of t h e a n t colonies on how they find t h e food source and bring them back to their nest by building the unique trail formation. T h e U C problem solved using the proposed approach is subject to real power balance, real power operating limits of generating units, spinning reserve, start u p cost, a n d minimum u p a n d down time constraints. T h e proposed approach determines t h e search space of multi-stage scheduling followed by considering the unit transition related constraints d u r i n g t h e process of state transition. T h e paper describes the proposed approach a n d presents test results on a IO-unit test system t h a t demonstrates its effectiveness in solving the U C problem.Index Terms -A n t colony search algorithm, distributed cooperative agents, optimization, unit commitment.

Adaptive and Dynamic Ant Colony Search Algorithm for Optimal Distribution Systems Reinforcement Strategy

“…In [3,4], the Authors use the ACO, Ant Colony Optimization, to solve the generation unit commitment problem. They have created a discrete search space, by considering a multi-stage scheduling and a suitable transition cost from one stage to the other.…”

A probabilistic framework for unit commitment problem in deregulated markets with high penetration of wind and solar power

“…On one hand, many classical optimization methods, such as Lagrangian relaxation (LR) [1], mixed-integer programming (MIP) [2], dynamic programming (DP) [3], linear programming (LP) [4], branch and bound (BB) [5], Benders decomposition (BD) [6], interior point optimization (IPO) [7], and priority list method [8], have been used for the solution of the deterministic UC problem. On the other hand, there are metaheuristic methods such as genetic algorithms [9,10], tabu search [11], evolutionary programming [12], simulated annealing (SA) and its hybrids [13][14][15], seeded memetic algorithm [16], and ant colony search algorithm [17] that can be used to solve this complicated problem. As noted earlier, the goal of the UC problem is usually cost minimization.…”

Mutation particle swarm optimization algorithm for solving the optimal operation model of thermal power plants

“…8,9 MIP guarantees convergence to the optimal solution in a finite number of steps while providing a flexible and accurate modeling framework. There are some smart algorithms to solve the UC problem, such as artificial neural networks (ANNs), 19 evolutionary programming (EP), 20 ant colony system algorithm (ACS), [21][22][23] genetic algorithm (GA), 24,25 binary coded GA (BCGA), 26 integer coded GA (ICGA), 26 genetic algorithm based on unit characteristic classification (UCCGA), 27 LR and GA (LRGA), 28 matrix real-coded GA (MRCGA), 29 floating point GA (FPGA), 30 mimetic algorithm (MA), 31 stochastic priority list (SPL), 32 extended priority list (EPL), 33 priority list based evolutionary algorithm (PLEA), 34 artificial fish swarm algorithm (AFSA), 35,36 particle swarm optimization (PSO) algorithm, 37,38 PSO combined with the LR (PSO-LR), 39 and improved PSO (IPSO). 38 Algorithms listed above have their respective advantage.…”