Research And Application Of Particle Swarm Optimization Based On Simulated Annealing Method In The Urban Layout Of The Land Space

Particle Swarm Optimization (PSO) is an optimization algorithm based on iteration, which was proposed by Kennedy and Eberhart in 1995. The system is initialized with a set of random solutions, and the best solution is to be found through a way of iteration. The algorithm compared with the genetic algorithms (GA), it is simple and easy to achieve, there is no crossover and mutation operations, the need to adjust the parameters is small, and it has fast convergence. Now it has been applied in the objective function optimization, dynamic optimization of the environment, the neural network training and many other fields, and in evolutionary computation IEEE Annual Meeting (IEEE Annual Conference of Evolutionary Computation, CEC) as an independent research branch.Co-evolution algorithm has been developed as a kind of new evolution algorithm extensively on the base of co-evolution theory for a decade.The difference between co-evolution algorithm and traditional evolution algorithm is that co-evolution algorithm takes the adjustment of populations as well as that of populations and environment into account in evolving process.For the advantages of co-evolution algorithm,more and more scholars contribute to this field,which makes it a hotspot in evolution computation.Simulated annealing algorithm was proposed by S. Kirkpatrick and others in 1982. It is a simulated annealing metal and the establishment of the mechanism of random optimization methods. It is a common random search technology which is based on direct simple simulation of solid annealing process, it is of a strong local search capabilities, and it can be able to avoid local optimal solution. This advantage is due to the existence of people thinking the successful introduction of combinatorial optimization theory, and in recent years the algorithm attracted wide attention of many areas, such as the design of large-scale optimization,numerical analysis, the complexity of the layout.Urbanization is a common phenomenon which exists in the today's developing countries, it shows the formation of new towns, the old city's expansion and demographic changes in agriculture and other non-agricultural population, so the preparation of urban land use planning programs has become an important task that planners often have to face. How to produce urban land use planning programs and how to evaluate and select the best program becomes one of the specific problems that they have to solve. For some areas, there are many viable urban planning programs, and to find each program is not only feasible, but also unrealistic. Currently, planners often can be only based on qualitative methods to select candidates for the program, the process of selecting is called a "black box" operation, the results are subjective, and the number of programs are very few. In practice, there may be a better program that was not found, so the search for an effective system is very necessary and urgent.Urban land use planning, that is the layout of urban land space (or space configuration), which belongs to the question of the allocation of resources, with a combination of challenges. At present, to solve such problems are of two categories: exact algorithm and heuristic algorithms. Exact algorithms include linear programming, integer programming from 0 to 1, and so on; heuristic algorithms include genetic algorithm (GA), simulated annealing algorithm (SA), Tabu search and artificial neural networks and so on. These algorithms are all related to the evolutionary algorithm.As urban land use planning is developing, evolutionary algorithm has become the most important planning method.This article will be focused on the combination of particle swarm optimization,co-evolutionary algorithms and simulated annealing algorithm, building a multi-particle swarm co-evolutionary model based on simulated annealing and applying it to the layout of urban land space.The main work of this article:1. Propose a multi-particle swarm co-evolution algorithm based on simulated annealing.Although traditional PSO is simple and easy to program, and has a good convergence performance, there arealso many shortcomings,such as it is easy to fall into local maximum points, has poor accuracy and slow convergence of the latter part of the evolution. In order to overcome these shortcomings, this article will combine simulated annealing with multi-particle swarm co-evolutionary,and propose a multi-particle swarm co-evolution algorithm based on simulated annealing. The simulation shows that the algorithm can not only enhance the particle swarm algorithm global convergence, but also to speed up the pace of evolution.2. Construct a multi-objective fitness function of the city's land planning.City land-use changes are a Complicated process which is of multi-variable, multi-object and is affected by various factors. As a result, land-use optimizing is a typical multi-objective decision-making, and considering a single goal as optimization of direction is clearly unreasonable. This article evaluated the situation of the city land-use from the coordination function of the land, the proportion of co-ordination of land, various types of urban land use in the spatial distribution of the overall co-ordination,population, transportation. According to it, the article analyzes the structure of land use distribution model and defines its objective function.3. Solve the city's land-use Optimizing program.This article uses the multi-particle swarm co-evolution algorithm based on simulated annealing to optimize the objective function of the layout of urban land space. In Windows XP,it generates a reasonable program with VC + +. NET 2003, and the results are satisfactory.