DocumentCode :
2466645
Title :
An Evolution Strategy for Improving the Design of Phased Array Transducers
Author :
Chen, Stephen ; Razzaqi, Sarah ; Lupien, Vincent
Author_Institution :
York Univ., Toronto
fYear :
0
fDate :
0-0 0
Firstpage :
2859
Lastpage :
2863
Abstract :
Continuum Probe Designertrade by Acoustic Ideas Inc. is a tool that can help design an optimized phased array ultrasonic transducer for a given inspection task. One component of Continuum Probe Designertrade can calculate the cost of a probe for a given geometry, and the fmincon function in the MATLABreg optimization toolbox has been used to find the optimal probe geometry. However, the search space of probe geometries is very noisy, and the inconsistent performance of fmincon in this noisy search space has interfered with the deployment of Continuum Probe Designertrade as a commercial product. An evolution strategy has been developed that includes "smart" start points and coordinated parallel runs. Compared to fmincon, the developed evolution strategy finds solutions that require an average of 61.4% fewer elements. Since control instrumentation can only handle a limited number of independent channels, the evolution strategy designs are also more than five times more likely to be implementable. These improvements that have been achieved by the evolution strategy will play an important role in the commercial deployment of Continuum Probe Designertrade.
Keywords :
CAD; evolutionary computation; inspection; mathematics computing; optimisation; production engineering computing; search problems; ultrasonic transducer arrays; Acoustic Ideas Inc; Continuum Probe Designer; MATLAB optimization toolbox; control instrumentation; evolution strategy; inspection task; optimal probe geometry; phased array ultrasonic transducer design; search space; Acoustic arrays; Acoustic transducers; Cost function; Design optimization; Geometry; Inspection; Phased arrays; Probes; Ultrasonic transducer arrays; Ultrasonic transducers;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Evolutionary Computation, 2006. CEC 2006. IEEE Congress on
Conference_Location :
Vancouver, BC
Print_ISBN :
0-7803-9487-9
Type :
conf
DOI :
10.1109/CEC.2006.1688668
Filename :
1688668
Link To Document :
بازگشت