Research On Project Multi-objective Optimazation Problem

Since the late of20th century, the construction industry and project managementtechnology developed rapidly. People are also increasingly concerned about the efficiency ofthe project. Multi-objective optimization techniques in the field of project management arewidely applied. Therefore, the project multi-objective optimization problem not only in thetheoretical research but also in practice is worth studying. However, most researchers beforestudied a single goal or two goals a linear constraint relations and solved using traditionalmathematical methods.This operation is more complex and the solution efficiency is alsolow, especially for engineering multi-objective optimization problem. The solution is notvery satisfactory. The particle swarm algorithm is relatively simple and convenient forengineering optimization.This paper discusses in detail the domestic and international multi-objectiveoptimization problem project development. Respectively from the traditional mathematicalsolution methods and modern intelligent optimization algorithms introduce the relevanttheoretical basis, analyzing the cost, quality, duration three goals connotation, projectduration and cost qualitative relationship. Combine the economics and the theory ofmarginal utility power function with parameters, taking into account the actual project inreward and punishment system. Construct a punitive system project duration-cost model.By studying the literature, summarizes the actual situation of the project and analyzes theduration and quality of the qualitative relationship, acquiring a relationship curve betweenthem. Process quality is quantized to0-1range to be judged and constructes the Projectduration-Quality Model on the basis of this. Then from the economy, quality assurance andtime integrate the project into an integrated multi-objective optimization model. On the PSOresearch, analyzes the content and MATLAB. Based on standard particle swarm algorithm,introduces the improved particle swarm algorithm, namely a compression factor weights andadaptive particle swarm particle swarm. Compared the three algorithms,adaptive weightparticle swarm is the best method in computing accuracy and convergence through two testfunction results.Then by drawing the crossover genetic algorithm, improves the adaptiveweight particle swarm and proposes the cross-factor PSO, appliying to practical problems.Finally, this paper introduces an engineering example and the crossover factor particleswarm optimization algorithm is applied to the project described in this article, so that theproblem can be enriched and demonstrate that the proposed particle swarm optimizationalgorithm is correct and reasonable. I hope the results of research in this paper can solve the actual problem and can beapplied and can improve the experience and demonstration of the project management forthe future.