The Problem Of Multi-projects Investment Based On Artificial Fish-Swarm Algorithm

In the era of rapid economic development, a lot of enterprises or companies make multi-investment projects have become the focus and hot. According to the limited resources (such as funds, equipments and human resources), it’s the core problem how to choose the best project portfolio, which can enable the enterprises to achieve maximum economic benefits, from so many alternative projects. At the same time, it is the key factor how to allocate the funds to the selected projects in reason, in order to resolve the conflict of funds under the environment of multi-investment projects. To solve the issue can effectively shorten the cycle of projects and greatly heighten the organizational effectiveness. Consequently, reasonable allocation of funds on each project has a significant relationship for obtaining greater benefits. So far, few scholars solve the problem of multi-investment projects based on intelligent optimization algorithms. Accordingly, it is attempted to solve such problem based on Artificial Fish-Swarm Algorithm in this article.Artificial Fish-Swarm Algorithm (which is abbreviated as AFSA) is a new kind of stochastic optimization algorithm based on the simulation of fish behavior. It has some characters, such as easy to programming, no special requirement for the initial values, the setting of parameters and object functions, stronger robustness and the ability of getting global convergence. However, with constantly study, some shortcomings of AFSA gradually exposed. Therefore, the improvement which is aimed at shortcomings and application of AFSA are studied by simulation. The main contents are as follows:Firstly, migrants are introduced to the basic AFSA for the purpose of contacting with multiple populations to search in parallel. Combined with the characteristics of basic AFSA, multiple-population AFSA is formed. The improved algorithm effectively compensates the deficiencies and improves the efficiency of searching.Secondly, a new hybrid optimization algorithm is produced by the combination of basic artificial fish-swam algorithm and artificial bee colony algorithm. The improved algorithm takes full advantage of the two kinds of intelligent algorithms to speed up the rate of getting the optimal solution.Lastly, a mathematical model is established for the selection of optimal portfolio and the allocation of funds in the multi-investment projects. The results of simulated experiments, which are based on basic AFSA, multiple-population AFSA and the hybrid optimization algorithm, show that they are efficiency to solve such problems. Meanwhile, the improved artificial fish-swarm algorithms are verified the superiority of solving multi-investment projects by comparing the results.In the thesis, the effective application of AFSA for multi-investment projects expands the areas of application of the algorithm.