| Now, the informationization is very popular in China, and the government takes it very seriously. The government not only sets up a special committee to guide the informationization process, but also put several ministries together to one ministry: the Ministry of Industry and Information Technology, which is ranked number one in China. It shows that the government’s attention and support. It is undeniable that enterprise informatization is core part and foundation of China’s Informatization. This is not only decided by the enterprise’s important position in market economy, but also by the special conditions of China. However, a severe challenge we have to face is: the high failure rate of the enterprise informatization project. So, the enterprise informatization investment has become a very popular, important and valuable research field. Many scholars have made outstanding contributions in this field.The topic of this thesis comes from a National Natural Science Fund called “The choice orientation and resource allocation for the informatization upgrade investment: based on the EICMMâ€. Moreover, this thesis was funded by the fund. The thesis researches the history and current condition of enterprise informatization in a systematic way, and summarizes the researching methods of enterprise informationization investment. Then, I decided to research bilevel programming model in enterprise informationization investment field. Bilevel programming is the main tool in this paper. The paper reviews the bilevel programming theory, and introduces the solution of bilevel programming model: chaos optimization algorithm. In this paper, according to the relevant theory and the actual situation, I build the informatization investment decision model with bilevel programming structure. And I got the chaotic optimization algorithm of this model by researching for a long time. This is the model building stage.By collecting a lot of investigation questionnaires, with the help of economics and management knowledge, I used the method of nonlinear regression to obtain a lot of benefit functions. These functions are classified by different subdivision industry of manufacturing industry and different scale. All of these functions can build up a function library. Based on this library, I decided to build a one-level programming model and get the optimum value first. According to the algorithm of one-level model, it is not very hard to get suitable algorithm for the bilevel model. Then I built a bilevel programming model with real data. In this model, the enterprise group is the upper layer and the sub companies are the lower levels. I achieved the realization of the algorithm by Matlab. At last, I got the optimal informatization investment scheme, while the group interest is regarded as the first goal. This process provides a relative and scientific decision-making method to support enterprise informatization investment decision in the real world. |