Research On Methods Of Time And Cost Equalization Based On N-level Period

ERP system has been used as an important tool of modern business management and demonstrates the growing importance, the main idea of which is to make allocation of resources enterprises owned rational, reduce production costs, improve productivity and make enterprises to acquire wealth maximize enabled. The high efficiency and low cost were seeking by companies in business and management of the project implementation process. So, how to reduce costs and improve efficiency has became an important problem for companies. And for this, the problem was discussed from the application of project in ERP system.The problem of higher efficiency and lower cost can be attributed to time-cost equalization problem of operations research. And now, some solutions have already been widely applied in business operations. After analysis, some deficiencies were found in solutions. So, the problem was considered from the perspective of local optimization, for which nonlinear model based N-level period was introduced, by which linear model was replaced. And then, the methods of time and cost equalization on the nonlinear model will be studied.In the research process of solutions, let the idea of mathematical programming applied on N-level period model. For this, the solution of mathematical programming is improved by making use of the characteristics of nonlinear model which is closer between time and cost than linear model. And the result acquired by estimates is more excellent. Secondly, the solution combined genetic algorithm with mathematical programming was proposed from the perspective of local optimum to resolve the problem. And finally, the solution for the characteristics of the local optimization and the deficiencies of combined genetic algorithm with mathematical programming, one solution which is based on local optimization and mathematical programming was proposed, by which not only shortcomings of the solution combined genetic algorithm with mathematical programming are avoided but also more simple in implementation, low time complexity and greatly reducing the computer calculation.After a theoretical research, the advantages and disadvantages for each solution were verified in experiment in the last part of this paper. From the experiment, some results were acquired that all of the solutions based on N-level period model in terms of estimates is better than the solutions based on linear model, and the best solution is the solution based on local optimization and mathematical programming.