A Study On The Application Of Artificial Intelligence In Direct Torque Control

With the China Telecom reintegrating, the competition of different companies related with the Mobile Services is serious, whereas,the difference of network service is reducing from time to time.As a result,competing rivals will have to surfer the loss due to merely the single price war.Almost all the telecommunications enterprise are looking for a good solution to improve their service quality as well as the competitive ability.These telecommunication enterprise have to improve the internal scientific decision-making ability and meanwhile strengthen the judgment ability urgently to face more severe competition in the market.So telecommunication enterprise need for a decision support system which from customer value analyse.Actually,the telecommunication enterprise have accumulated a mass of digitalized call information.They can search out a lot of valuable information from call data with the aid of data mining techniques,take customer consumption characteristics for example.Base on customer consumption characteristics we can assessment of customer value and doing the customer value segmentation.Such data can be further interpreted by the marking department to provide more powerful marketing strategy and also reduce a certain cost.Taking the customers of Zhejiang Mobile Communication Company as research object,the evaluation system of customer value as well as customer value matrix is established in this paper.Dividing the customers into low-value customer,secondary-value customer,potential-value customer and high-value customer those four segments according to the CRM and characteristic of business.In the process of the study,on data pre-processing with hierarchical encoding method.The method has many advantage,such as safe and easy expansion and so on.The customer value consists of two parts:one is current realized value;the other is future potential value.Using frequency statistics and Least Square Method,to get the critical value ,the critical value model of the current realized value is founded.The multiple clustering method and ID3 are further introduced to build Credit model and Royalty model so that the evaluation of the customer future potential value can be achieved.Compared with traditional hierarchical clustering method ,multiple clustering method greatly enhanced accuracy of clustering results with increase the clustering of index.Then customer value matrix is founded by means of customer value ,then enterprise to do marketing strategy of customer value promotion. In this paper,we mining Zhejiang Mobile Communication Company users' call information to identify relationship customer.Through the research,those companies can find the deep requirements of the customer and design a more reasonable combination.Base on case study in real applications,the author has further proved the validity and effectiveness of this model.The research results can be taken as direct guidance for development of customer value management of Zhejing Mobile Communication Company., and canalso be referenced by other teltcommunication companies.