Design And Implementation Of Assistant Decision-making Subsystem Of Hf Frequency Selection And Communication System

The content of this paper is derived from the laboratory's project,HF Frequency Selection and Communication System,which aims to realize HF communication at sea.The system contains functions of link establishment,data transmission,frequency selection and frequency update.This paper innovates in the topic selection,adding assistant decision-making subsystem to the system,which makes up for the deficiency that the system has not yet had the assistant decision-making function.This subsystem can improve the efficiency of the system,reduce the blindness of user operation and improve the friendliness of equipment operation by providing users with operation suggestions and assisting users to make operation decisions.The main tasks completed are as follows:1.The main functions and implementation principles of the key modules(link establishment module,data transmission module,frequency selection module and frequency update module)of HF Frequency Selection and Communication System are analyzed,so as to prepare for the requirement analysis of the assistant decision-making subsystem.2.The functions of the assistant decision-making subsystem mainly be divided into the following five aspects: detection frequency selection advice for blind connection,troubleshooting advice,selection of methods for link build or frequency update,frequency selection recommendation and report generation.3.In order to realize the assistant decision-making subsystem,build the database,knowledge base,method library and model library over the HF Frequency Selection and Communication System.The database is implemented based on My SQL software,which is used to store the data information generated during the communication of the system and provide data support for decision-making.The knowledge base is established by the production system,which is used to store the knowledge and suggestions of experts and provide the basis for qualitative decision.In order to realize various functions of the subsystem,the frequency connectivity model,link building/ frequency update method mode selection model,troubleshooting model and frequency comprehensive evaluation model are established,and these models are put into the model library.In the method library,mathematical methods such as data filtering method and improved combination weighting method used in the model library are stored.These methods are stored in the method library in the form of function modules for the model library to call.4.An innovation is presented during the realization of frequency selective suggest function of assistant decision-making subsystem,which is a frequency comprehensive evaluation model based on improved combination weighting method.Predicting the frequency with real historical data make up the deficiency of neural network method's lack of training a large number of real data,and the deficiency of Seven-day weighted average method's simple.5.The programming realization and indoor test verification of the assistant decision-making subsystem were carried out.The experimental results show that the subsystem can give reasonable suggestions and the operation is stable and reliable.