Research On Multi-objective Optimization Of Engineering Projects Based On Improved Particle Swarm Algorithm

In recent years,China’s construction industry has been rapidly developing,bringing along a series of challenges.In the field of project management,time,cost,and quality are three crucial aspects.Therefore,the problem of balancing time,cost,and quality(TCQT)has always been a significant challenge in the construction industry,as it involves balancing multiple optimization objectives.Traditional construction planning methods often rely on historical data and the experience of managers,which may not provide sufficient reliability and accuracy.Particle Swarm Optimization(PSO)is one of the introduced mature metaheuristic algorithms.Although metaheuristic algorithms have been proven to possess superior characteristics compared to other traditional methods,they also have limitations.For instance,they can easily get trapped in local minima or converge too slowly when dealing with single-peaked problems.Existing research on addressing the issue of algorithm’s susceptibility to local optima is also limited.Furthermore,the majority of existing studies focus on bi-objective linear or nonlinear research concerning time-cost and time-quality,with limited consideration given to multi-objective optimization involving factors such as quality or safety risks.It is worth noting that the basic Particle Swarm Optimization algorithm is based on searching in continuous space(interval).However,in some practical engineering applications,the variables to be solved may not be continuous but discrete in nature,such as different resource allocation plans or construction methods.Moreover,different resource allocation plans or construction methods correspond to different resource allocations,including equipment,personnel,and materials,which also exhibit different performance in terms of time,cost,and quality.Existing research on TCQT problems typically adopts precise numbers to describe time,cost,and quality.However,due to uncertain environments and subjectivity,it is difficult to use precise numbers to describe these performances accurately.Finally,there is a lack of research on handling Pareto optimal solution decisions.This paper addresses the shortcomings of previous research and analyzes the following aspects.Firstly,based on the standard multi-objective Particle Swarm Optimization algorithm,Cat chaos and Gaussian mutation improvement strategies are introduced and tested using three single-objective functions and three multi-objective function test functions to demonstrate the superior performance of the improved multi-objective Particle Swarm Optimization algorithm in obtaining more accurate Pareto optimal solutions.Secondly,based on the same engineering case,a multi-objective continuous optimization and discrete optimization model for time-cost-quality is established,and the fuzzy multi-attribute utility theory is introduced into the Particle Swarm Optimization discrete optimization model to describe the fuzzy,uncertain,and imprecise performances related to resource allocation plans or construction methods,aiming to search for the optimal combination of construction strategies.Finally,a combination of weighting and grey relational analysis is proposed to rank multiple sets of Pareto optimal solutions in the continuous optimization model,and a comparison is made between the optimization results of the two modes,along with an analysis of their differences.