Research On Application Of Improved Water Cycle Algorithm In Multi-objective Active Power Optimization Problem

Active power optimization is a very important link in the economic dispatch of the power system.With the continuous increaseing in power demand and the more complex power grid system,researchers have focused more attention on the study of Multi-objective Optimal Active Power Dispatch(MOAPD),the targets involved basic fuel costs,power loss,emissions of harmful gases such as nitrogen sulfide,and consideration of fuel costs with valve point effects.Because different objective functions usually have a competitive relationship and have different dimensions,the difficulty of solving the MOAPD problem is greatly increased.This thesis chooses the water cycle algorithm(WCA)as the main solution to solve the MOAPD problem.Firstly,in view of the deficiencies of the standard water cycle algorithm in solving the non-convex optimization power flow problem,the Gaussian mutation mechanism and the water flow evaporation process are introduced to improve the basic WCA,and a novel Gaussian Water Cycle Algorithm(NGWCA)is proposed.And apply the algorithm to solve single-target simulation experiments.Furthermore,on the premise of verifying the superiority of NGWCA in single-objective simulation,the sorting idea in the NSGA-III algorithm is borrowed,and multi-objective strategies such as global sorting rules containing rank and distance attributes and elite solution set retention strategies are incorporated into NGWCA.The algorithm is changed to an algorithm that can solve multi-objective problems.In addition,the Constraint Dominant Strategy(CDS)is proposed,which improves the original constraint processing rules,avoids the complicated process of unifying dimensions and a reasonable range of adjustment coefficients.Combining the multi-objective NGWCA and CDS dominance strategies,NGWCA-CDS algorithm is proposed.In order to verify the effectiveness of NGWCA-CDS in solving the MOAPD problem,the NGWCA algorithm is compared to the Cuckoo Search Algorithm(CS),Particle Swarm Optimization(PSO),and Differential Evolution(DE)on the IEEE 30 and IEEE 57 test systems.It is applied to solve the MOAPD problem,simulates 6dual-objective and 1 three-objective simulation experiments,and tries to apply the algorithm to the optimization of the IEEE 118 large-node system.The simulation results are compared with MOPSO and NSGA-III.The results show that the improved method can effectively solve the MOAPD problem.Under the premise of fully satisfying the constraints,not only a consistent and continuous Pareto solution set can be obtained,but also a competitive compromise solution can be obtained.In addition,from the perspective of generation distance,distribution,and computational complexity,the evaluation indicators also show that the NGWCA-CDS algorithm is evenly distributed,has high convergence and strong stability.Further evidence shows that the algorithm proposed by NGWCA-CDS can effectively solve the MOAPD problem.