Sensor false alarms

PRINCIPLE CHEMICAL SOURCES OF SENSOR FALSE ALARMS... 

The selectivity of previous Fido sensor prototypes for landmine chemical signature constituents has proven to be excellent. However, improvements in chemical selectivity for target analytes translates into better false alarm performance, so methods for reducing sensor false alarms have been a focal point of sensor research efforts. Previous sensor prototypes have achieved selectivity primarily through the chemical and physical properties of the amplifying fluorescent polymer (AFP) materials used as sensoiy materials in the system. Described here are changes to the sensor hardware that should further enhance the selectivity of the sensor. 

Vibration sensors are similar to break-glass detectors but are fixed to walls and doors against violent attack. The stmcture to which they are fitted must not be subject to vibration from external sources, otherwise false alarms may result. 

Reliability. Sensors must indicate the presence of hydrogen specifically, and not provide false alarms. The sensors should have consistent reproducibility (e.g., 5% over 1 year at 2% v/v H2 in air), long-term stability and minimal drift rates. Response must not drift outside acceptable limits over that lifetime without providing an alarm. Figure 15.3 [3] shows an example comparison of the theoretical versus measured... 

Fig. 5.42 Power Activation Mode -operation sequence for Ga203-natural gas sensors to avoid false alarm in the presence of ethanol and other interfering air components [20], (With kind permission of D. Kohl, D. Skiera and M. Lam-mer, Institute of Applied Physics, University of Giessen, Germany)...

Multi-sensor fire detectors provide faster detection with fewer false alarms. These sensors are capable of monitoring the environment for multiple purposes, e.g. carbon monoxide concentration, concentration of flammable gases and indoor air quality, too. 

In addition to sensitivity, an important feature for any biochemical sensor is selectivity, i.e., the ability to response to a specific chemical or agent. In additional to the obvious significance (reducing the false alarm rate, etc.), this ability is... 

Keywords wireless sensor network detection theory Kalman filtering target intrusion detection false alarm. 

The field to be monitored is usually narrow and long in perimeter security applications. Thus, non-uniform deployment may be necessary. He et al. conclude that the sensor nodes generate false alarms at a non-negligible rate [18], and an exponentially weighted moving average on the sensor node is sufficient to eliminate transient alarms. 

The model and results developed herein give clues that link false alarms to energy efficiency. Enforcing a low false alarm rate to avoid unnecessary response costs implies either a larger data-set (L) and hence a greater battery consumption, or a denser sensor network, which increases the deployment cost. Similar qualitative and/or quantitative inferences about the relationships between various other parameters can also be made. 

In general, the tracking procedure starts with an association process to combine the established track parameter with the radar sensor or radar network measurements. Errors in the association process will always lead to ghost targets. But the general requirement for automotive applications is to keep the false alarm probability as low as possible, which underlines the importance of the association process for radar networks. 

A global diagnosis system (i.e., accounting for all the available information sources in a single box) could have been developed. However, modularity would then be very low since adding or removing one information source (e.., one sensor) would break down the overall structure and lead to false alarms or... 

Figures 6.5 and 6.6 are referred to a slow drift of the output of sensor Sj, i.e., a linearly increasing signal, with a 10-3 K s 1 rate of change, is added to the measured variable for t >t = 9000 s. It can be recognized that the fault is detected a few time instants after the occurrence, while it is isolated about 2000 s after tf. This is due to the slow time evolution of the fault it can be argued that, in the first 2000 s after the occurrence of the fault, its effect is quite negligible and/or almost totally compensated by the observers. In order to reduce the isolation time, the normalization factors could be reduced, at the expense of an increased probability of false alarms. Moreover, Fig. 6.5 shows that the voted measure is the mean value of the measured and estimated data until the isolation is performed, and then it switches to the value of the healthy sensor (5), i).

---