الگوریتم هیوریستیک برای حل مشکل واحد تعهد برای زندگی واقعی سیستم های قدرت در مقیاس بزرگ

One of the main needs that power system operators around the world have is to solve complex Unit Commitment models for large-scale power systems in an acceptable computation time. This Paper presents an alternative Heuristic algorithm that successfully addresses this need. The Heuristic algorithm makes use of various optimization techniques such as Mixed Integer Linear Programming (MILP), Quadratic Programming (QP), Quadratically Constrained Programming (QCP), and Dynamic Programming (DP). CPLEX 12.2 is used as the main optimization engine for MILP, QP, and QCP. DP is an in-house algorithm used to obtain the commitment of Combined Cycle Plants (CCPs) when represented with the component-based model. This Heuristic algorithm combines the global optimality capabilities of MI (L) P formulations with the highly detailed models available for CCPs using LR–DP formulations. The Heuristic algorithm introduced in this Paper is capable of solving up to 1-week scenarios with a 1-hour time window for the complex Mexican Power System.

Unit commitment problem is one of the most difficult hard limit optimization problems which are affected by some especial constraints that are imposed from system and physical conditions. Solving the UC problem is important from both the execution time and the correct lay out of plants with minimum cost aspects. To Many text resources have been published in the field of UC problem. Below there is a brief look at the UC problem solution methods in the recent literatures. The priority list (PL) [1]–[2] commits in ascending order of units with full-load cost so that the most economic base load units are committed first in order to meet the load demand. The PL method is very fast but highly heuristic and gives schedules with a relatively higher operation cost. The branch-and-bound (BB) method [3]–[4] has the danger of a deficiency in storage capacity and increasing the calculation time enormously encountering with large-scale problem. The Lagrangian Relaxation (LR) method [5]–[7] concentrates on finding an appropriate coordination technique for generating a feasible primal solution while minimizing the duality gap. The main problem with the LR method is the difficulty encountered in obtaining feasible solutions. The meta-heuristic methods are iterative search techniques that can search not only local optimal solutions but also a global optimal solution. In the meta-heuristic methods, the GA, TS, EP, SA and etc are used for UC [8]–[11]. These methods have the advantage of searching the solution space more thoroughly and avoiding premature convergence to local minima. The main difficulty is their sensitivity to the choice of parameters. However, in case of a large-scale problem, they consume a lot of time and space due to their iterative nature.

In this paper, a novel optimization algorithm based on Harmony Search (HS) is adapted and applied to UC problem. The proposed method is first implemented in ten units system and the obtained results are compared with the results given in the references. At the second step of the study, the test system is modified and duplicated for 20-100 units system. The obtained results show that the proposed method can solve both small scale and either the large scale UC problems effectively. The calculatedcostand execution time is compared to the other well known optimization methods. Based ontheobtained results, the algorithm is an effective method to solve the complex,nonlinearandhardsatisfactoryoptimization problems.