An Approach For Solving Short-term Hydrothermal Scheduling Based On Gravitational Search Algorithm

The short-term hydrothermal optimal scheduling has huge economic benefits inelectric power system operation, so it is always a hot issue for researchers. In the powersystem scheduling problem, the traditional mathematical model only considersmaximizing the economic income, the objective of the problem is making full use of thehydraulic resource and minimizing the fuel cost of thermal plants. With the rapiddevelopment of modern society, the energy shortage and environment problem becomeserious, energy efficiency and environmental protection are the nevitable choice for thestrategy of sustainable development. It is significant to take the emission as one of theobjectives of hydrothermal scheduling problem. Researchers have proposed manyoptimization stochastic search algorithms for solving hydrothermal scheduling problem,but most of these algorithms suffer from premature convergence, and the strategy ofconstraints handling is rarely developed. This paper improves the search performance ofthe gravitational search algorithm, and proposes the multi-objective gravitational searchalgorithm to solve short-term economic/environmental hydrothermal scheduling. Themain works of this paper are described as follows:1) In order to overcome the drawback of the premature convergence, the paperproposes an improved gravitational search algorithm (IGSA). Firstly, the paper introducesparticle memory character and population social information in velocity update process.And a chaotic mutation is adopted to enlarge the search direction from the region near thelocal optimal to the global feasible region, which improves the performance to find theglobal optimal solution. Furthermore, IGSA utilizes a rule based on selection operator forpopulation evolution; it ensures the population always evolves towards the global optimalsolution. To deal with the multi-objective optimization problem, by introducing theconcept of non-dominated sorting and crowding distance, this paper develops anon-dominated sorting gravitational search algorithm with chaotic mutation. The tests ofbenchmark problems prove that NSGSA-CM has good search ability and it is efficient tosolve the high dimension optimization problems.2) In the background of energy efficiency and environmental protection, this paper formulates the mathematical model of short-term economic/environmental hydrothermalscheduling problem. It is a bi-objective optimization problem which takes both fuel costand emission of hydrothermal system into account. And by introducing the variableweights based on time interval, the bi-objective problem is converted to a single one andbecomes suitable for IGSA.3) When dealing the various constraints of the problem, this paper applies heuristicadjustment strategies. A modification strategy by dividing the violation water volume intoseveral parts and randomly selecting intervals to adjust the water discharge gradually isproposed to handle the water dynamic balance constraints. A new strategy is adopted tomeet the reservoir storage volumes limit constraints. Based on the violations of thereservoir storages, it adjusts the water discharge of the hydro plant at previous timeintervals in feasible region. Meanwhile, it modifies the equal amount to the correspondingdischarge at the later time intervals to keep water dynamic balance. A strategy based onadjusting the output of the thermal power plants according to their priority index is alsoapplied to handle system load balance constraints effectively.4) To verify the IGSA and NSGSA-CM are feasible and effective for solvingshort-term hydrothermal scheduling problem, this paper applies two test systems forsimulation computation. The results show that, IGSA and NSGSA-CM have betterperformance for searching global optimization in comparison with the approaches inliteratures, and they obtain less fuel cost and smaller pollution emission with satisfying allthe constraints. It is found that the algorithms and the constraints handing methods in thispaper are effective and efficient for dealing with short-term hydrothermal schedulingproblem.