The Design And Development Of Pulse Identification System

The skill of pulse condition identification in Chinese medicine has become a treasure of Chinese culture,which is an inevitable method for Chinese doctor to diagnose a disease.Simultaneously,the qualitative and subjective of this method can affect the scientificity of diagnosing a disease.Also,the result of this method can not be recorded and conserved,which is a limit to the intensive research of pulse mechanism.It is of great importance for the related researchers to give the pulse an objective description and impersonality diagnose.This thesis designed and developed the pulse identification system.Firstly,the general design and database correlation chart design has been fulfilled,and according to its functional configuration this system has been separated into pulse wave data acquisition module,pulse condition signal eigenvalue extraction module,pulse condition identification module,users' physiological parameter report module,encyclopedia of traditional Chinese medicine module and physical data module.Among all the modules,pulse condition identification module is the key research point.The research of pulse condition identification module contains two parts:(1)An new algorithm(APVD)has been put forward according to the mechanism of pulse condition signal.The test with actual data proved that the APVD algorithm can accurately abstract eigenvalues of every pulse wave from varieties complicated pulse condition signals.(2)On the basis of the characteristic of Chinese medicine,the ELM classification arithmetic has been put forward.This algorithm using KPCA to dimensionality reduction the eigenvalues and based on the front-feedback ELM input eigenvectors to identify common pulse.More than one hundred thousand pulse wave from volunteers has been tested and arrive at a conclusion:ELM has an at least seek-accuracy rate of 95.13%.In the end,by using SerialPort from NET Framework 4.0,the pulse wave data transmission has been realized.By using APVD algorithm the eigenvalues has been abstracted and then realized the identification of pulse.Moreover,by using ADO.NET database skills,stored procedure and the transaction model that data is properly operated and managed.The pulse identification system in this thesis is strictly according to the procedures of software development.The structure of hardware and software in deployed system and hardware platform are all used HKG-07B IR pulse wave sensor.The development tool is Visual Studio 2012 while backstage supporter database server used SQL Server2008,the C/S three layer structure,as a home edition software,this system realized the function of collect physical pulse data,analyzing and calculating,identifying pulse condition signal by eigenvalue,and finally give consumer a healthy assessment.At last,test for the system function.After the test,this system basically can satisfy the users' functional needs and non-functional needs.This system can get rid of doctors' objective emotion to the best advantage,making pulse diagnose more scientifically,thus bring this method's characteristic,such as economic,convenience,safe,no side effects into play,thereby making pulse skill has higher practical value.