Military training exercises produce a number of high-magnitude, short duration noise events, or impulse noise, that can may be a concern in nearby residential or environmentally sensitive areas. Noise monitoring systems are needed to quantify the magnitude and time of these impulsive noise events, to ensure compliance and to provide an archival record of noise emanating from the installation. Several commercial monitoring systems are available; however they are known to report excessive false positives as a result of windborne noise and distant non-military acoustic events. This result can bias noise statistics to the point where meaningful assessment of the acoustic sound levels from a site is not possible. Therefore, the Department of Defense (DoD) requires improved measurement and reporting of impulsive acoustic noise at military installations to address these concerns and to provide accurate noise information required by test and training managers, environmental safety/compliance officers and the general public residing adjacent to military installations.
APS, in collaboration with the University of Pittsburgh, have developed an improved noise monitoring system, called BAMAS (Bearing and Amplitude Measurement and Analysis System), for mitigating windborne and other sources of non-military noise. The objective of this project was to develop an improved system for the measurement and analysis of impulsive acoustic noise as well as the automatic detection and reporting of noise events.
Research and development of the BAMAS system was accomplished under a two phase effort. Phase I was a proof-of-concept effort focused on preliminary measurements of impulse noise and algorithm design. As part of this proof-of-concept phase, the requirements for individual microphones and microphone arrays were developed, and algorithms for impulse noise bearing and range determination and classification built. The multi-microphone array was then assembled and tested. The Go/No-Go decision for the BAMAS system was based upon whether or not the algorithms developed in Phase I could sufficiently reject events generated by windborne noise. Conclusions drawn from this experiment show that a small array of four (4) microphones can improve the false alarm rate (FAR) of existing systems (i.e., single microphone with simple detection processing) by a factor of 1000. This result led to a positive Go/No-Go decision.
In the second phase of this project, 2 BAMAS noise monitors were designed, fabricated, and tested. These units were installed near existing noise monitors at an active military base and the accuracy of the existing noise monitors was compared to that of the BAMAS noise monitors.
The techniques developed through this research have been tested and integrated into the APS BAMAS impulse noise monitoring system. Capabilities include detecting, archiving, and reporting all military impulse noise; (2) rejecting false sources of non-military noise from wind and aircraft; (3) measuring direction-of-arrival (DOA) information and providing source triangulation; and (4) developing improved software for user display and system maintenance. The BAMAS algorithm was found to reject 99.5% of the non-blast noise recorded (specifically wind) while retaining 97.7% of all blast noise. The system as a whole (sensors, base station, and software) functioned properly during field testing and is actively reporting new blast data to our online website/database.
The benefits of the new measurement system include more accurate detection and logging of impulsive noise events through rejection of wind-induced impulses, reduced user analysis time, better situational awareness through measurement of the bearing and elevation angle of the incoming blast, and automated impulse classification. Moreover, in the case that two noise monitors detect the same blast noise, this information can be used to estimate the latitude/longitude of the source. Users logged-in to the online tool may view blast noise in near real time using an interactive spreadsheet and Google Earth mapping program. The user can also display noise events in tabular or graphical form. Reports can be generated automatically as defined by the user and will include relevant statistics, such as the number of events per week or other time frame as well as maximum and root-mean-square event magnitudes.
A prototype of this system was installed at an active military base where it has been reporting and archiving impulse noise since October of 2009. In the first 2 months of operation, the BAMAS system archived over 3000 events, none of which none were incorrectly classified as blast noise during periods of significant wind.