Differential Evolution Algorithm And Its Application In Enterprise Informatization

With the speeding up of the Informatization process in nowadays, our country’s enterprise Informatization has been deepened. Enterprise informatization investment also involve the investment cost of hardware, software and other facilities in the enterprise investment share and finally get the more rational benefit of it. In order to allow enterprises to better developing, the enterprise in the information cost has been increasing, more and more enterprisers are considering the investment decision problems in the process of informatization. Although this problem has already become enterprisers’growing concerned problem, but now the enterprisers’research of informationization is still insufficient. Enterprise information system investment become improving only when the information systems need to upgrade. So if the investment become the actual prediction, that’s to say, most companies will make a reasonable forecast for enterprise information construction, this will do good to enterprise’s operation and management, which will also make a very good guidance function. Investment decision problems, and the information system is a function model of planning, can be summed up in function optimization problems. In research on the issue of enterprise information system investment, we can use economics related model. Differential evolutionary algorithm is a classic model of solving the questions of the enterprise information in recent years, its research on the enterprise Informatization problems are very widely used. The differential evolution algorithm has its unique advantages to solve the different problems in different fields. The characteristics of the differential evolution algorithm is simple easy to mix its beneficial to solve function optimization problems.This article is in the research of 《enterprise’ Informationization upgrade investment decisions》 the study of "Chinese national natural science foundation", ues the Matlab to simulate t differential evolution algorithm, test the population size and mutation factors. The range of parameters such as crossover probability fitness were analyzed, and the traditional differential evolution algorithm was improved, the traditional single fixed differential evolution factors such as variation factor interleaving, improvement for the adaptive differential evolution operator, based on regression of crowded initialization which make the initial population to better dispersion. Through the two improvement way, the simulation test proves that the improving not only speed up the convergence speed of the algorithm, but also improved the precision of the optimal solution. After improved differential evolution algorithm,then further introduced the double population mechanism, the first population using differential evolution algorithm and the group is used as a superior species, another population using evolutionary genetic algorithm (ga), the group is signed as inferior species. Through the selection mechanism,evolutionary game theory will be introduced to the inferior population to value superior populations.Through Matlab simulation experiment, the improved algorithm comparing with the traditional method to analyze its evolutionary speed, maximum, minimum, running time and so on. Algorithm is improved, the improved algorithm is applied in enterprise Informationization upgrade decision model, solve the practical problems of the process of enterprise information updating. Finally, I wish to turn the improved differential evolution algorithm on more and more enterprise information model, to solve function optimization problems.