In view such issues of the noise detection system of present UHV substation as high power consumption

high cost and cumbersome operation

a new noise detection system for substation is developed with combination ofMEMS sensor technology. The voiceprint characteristics and environmental characteristics of the substation noise sig-nal are taken as the base

the microphone elements are selected and encapsulated. The noise detection device has abuilt-in CORTEX M0 controller to calculate the A-weighted sound pressure level of the machine operating noise inreal time which is simultaneously sent to the the monitoring platform of host computer.The device can perform suchtwo kinds of acquisition commands as parallel data timing and threshold triggering for storing raw data.The whole de-tection system can achieve large-scale networking monitoring through Zig Bee

and the host computer monitoring plat-form sends instructions to complete the management and control of all noise detecting devices in the station. It isshown by the application results at certain substation that the detection system has high real-time performance andaccuracy

and its such features as low power consumption

low cost

concise and efficiency have great advantages inthe trend of UHV substation noise monitoring wide-area deployment under the construction of the internet of things.