Improvements And Applications Of Invasive Weed Optimization Algorithm

Invasive weed optimization (IWO) algorithm, is a novel intelligentoptimization model, proposed by Mehrabian and Lucas in2006, which is inspiredby a common phenomenon in agriculture: colonization of invasive weeds. Not onlyit has the robustness, but also it is easy to understand and program. So far, it hasbeen applied in many engineering fields. However, the drawback of being trappedin local optimum and lower optimization accuracy has greatly limited theapplication range of IWO. So, this paper will conduct the deep research to theinvasive weed optimization algorithm with the purpose of perfecting the basicIWO algorithm and expanding the application domain of IWO.The research results obtained are as following:(1)Aiming at the phenomenon of low accuracy and premature convergence ofinvasive weed optimization algorithm, this paper proposes a new invasive weedoptimization based on complex method, which employ the population diversity ofIWO and local research of complex method with relfex、extend and compress, soas to improve the optimization precision and convergence speed, meanwhile makethe algorithm effectively jump out of local minima.(2)According to traditional numerical methods are sensitive to initial iterationvalue, this paper proposed a invasive weed optimization algorithm which haspopulation diversity and combins with the heuristic global search of differentialevolution algorithm to solve Nonlinear Systems of Equation.(3)A method based on Invasive Weed Optimization Algorithm (IWO) withinequality point segmentation is presented to solve numerical integral so as toexpand the application domain of IWO.(4)An Invasive Weed Optimization (IWO) scheduling algorithm with the weed individuals be coede is presented for solving Flow-shop Scheduling Problem.(5) A Discrete I WO (DIWO) is proposed to solve the Traveling SalesmanProblem (TSP).