The Counting Improvement Of The Problem Of Flow Optimization In System Of Distribution Under The Form Of SaaS

Nowadays, many companies in our country often face the problem of high-paid atthe purchasing and product supporting in the process of distribution. The new anddeveloping SaaS will reduce the mentioned fee for the company by changing thescattering purchasing form into a concentrating renting form. There are manydifferent distribution system of the SaaS form in the market, and in order to make theSaaS form accept by the users in a short time, we should take the users operating habitand the operating procedure into consideration. This article will study from the users’operation habit of the computer and try to construct a corresponding screen to theusers and also introduce the idea of flow optimization.In the existing system,there has been a algorithm about the optimization of flow.This thinking is mainly derived from the statistical, use the history users’ actions toestimate the today’s users’ action. Through observation and analysis of existingalgorithms,we find some problems about the algorithm:1)the influence about do notdistinguish the impact of time for history users,2)Changes in the externalenvironment may cause unstable system performance. These problems will causenagtive affect to the system. Therefore,the motivation of this research is to study andanalyze the problems existing algorithms and to find a new avenue for settlement, andconduct experiments to verify.In order to solve problems in existing algorithms, this paper introduces the antcolony algorithm,it is a commonly used algorithm to find the optimal solution. In theant colony algorithm, the optimal solution may not be fixed,it will change.Because ofother features of ant colony algorithm is also consistent with the background of theproject,we use this kind of algorithm into our system and try to improve thealgorithm so that it can improve system performance.In this paper,by the experimental approach,we simulate the operation of the systembehavior by a user.It proved that existing method did have inadequates. Through awide range of experiments,simulate the various situations and compare the existingalgoritm with the improved algorithm based on ant colony algorithm,we find that it ispossible to improve system performance by use the improved algorithm to replace theexisting algorithms.It means thatcreate a custom user action in line with the operationof the desktop so that The economic benefits of the system can be improved.