Improved Neighborhood Field Optimization Algorithm And Its Application In Reinforcement Automatic Arrangement

Optimization problems are ubiquitous in real life,many researchers have proposed evolutionary algorithms to solve these issues.As a relatively new evolutionary algorithm,neighborhood field optimization algorithm has been applied in many fields.Different from the optimization mechanism of other algorithms,it utilizes local cooperation behaviors to explore the best solution to the problem.Experimental results in related literatures show that the transfer and exchange of local information will accelerate the convergence rate and search process.Therefore,the research of neighborhood field optimization algorithm based on local information for global search is of great significance for the development of intelligent evolutionary algorithm.In addition,due to the development of interdisciplinary,and considering the demand of intelligent and automatic development in the construction industry,this paper makes relevant application research on the problem of steel reinforcement automatic arrangement in the construction industry.In this paper,based on the introduction of the basic principle of neighborhood field optimization algorithm,the improved algorithm is proposed according to the accuracy of local information transmission in localization operation,and according to the requirement of steel reinforcement automatic arrangement in construction industry,neighborhood field optimization algorithm is improved to realize steel reinforcement automatic arrangement.The main research contents of this paper are as follows:(1)This paper presents a structured neighborhood field optimization algorithm based on local information refinement in localization operation.This method allows the individual to quickly collect effective local information,easily escape from local extremum,and quickly converge to the global optimal solution.The experiments on seven benchmark functions and CEC2014 standard test dataset show that the performance of the structured algorithm is superior to other algorithms.(2)For neighborhood field optimization algorithm and other algorithms,it is difficult to quantify benefits of local cooperation in the optimization process.For this purpose,this paper proposes a new network approach to analyze cooperation behavior.The population structure of structured neighborhood field optimization algorithm is studied and its structure is a scale-free network with power law distribution.Network characteristics,i.e.,cluster coefficient and average degree,are adopted to quantify the cooperation behaviors.Experimental results show that network characteristics can effectively transmit the optimization performance of structured algorithm in terms of convergence rate and population diversity.The algorithm with large clustering coefficient and strong heterogeneity degree distribution can solve the problem more effectively.(3)Finally,based on the mathematical model and industry design codes of steel reinforcement automatic arrangement problem in construction industry,this paper proposes a receding-horizon neighborhood field optimization algorithm based on large scale,multi-constraints and discrete minimization path problem,which is used to realize automatically collision and congestion free layout of rebar in the beam-column joint area of reinforced concrete building frames.In addition,for beam-column joints of three types(T type,+ type and L type,the architectural terms are Exterior joint,Interior joint and Corner joint respectively),three algorithms,namely,particle swarm optimization,differential evolution and neighborhood field optimization,will obtain optimum path for each reinforcement.Compared with the running time and the path length,the experimental results show that the result of automatic optimization meets the requirement of reinforcement layout,and neighborhood field optimization algorithm has the shortest placement time.