曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
would differ from region to region when expressed as a yearly chance of an accident. The
last two rows of the table demonstrate how the yearly chance of an accident may change
over time in the same region for a constant risk per flight hour. Conversely, Table 3-2
shows how the exposure to collision risk per flight hour would differ by region for the
same yearly chance of collision, based on 1995 traffic estimates. The figures indicate that
if the same yearly chance of collision is maintained from region to region, individuals
flying in North Atlantic (NAT) Minimum Navigation Performance Standards (MNPS)
airspace will, on average, experience roughly a seven times greater risk of en route
collision per hour than will individuals flying in European Civil Aviation Conference
(ECAC) airspace; while people flying in the Tasman area will, on average, experience
roughly 189 times greater risk than those in ECAC.
For this reason, the sole use of metrics such as the probability of an accident per year are
not advisable for expressing collision risk, as they give no indication, by themselves, of the
individual’s exposure to the risk.
The most commonly used metric for collision risk is “fatal accidents per flight hour.” This
metric is useful because it enables an individual’s exposure to risk to be linked directly to
the time spent flying, and it is not influenced by changes in such things as the amount of air
traffic or passenger levels. Another risk metric applicable to the individual is fatalities per
passenger trip. However, data needed to compute this metric are not as readily available.
However, although the metric of fatal accidents is arguably one of the most appropriate
for expressing collision risk, examining the risk expressed in other metrics is advisable and,
given sufficient data, is relatively straightforward. For example, if the risk in terms of fatal
accidents per flight hour, and the number of flight hours per year, are both known for a
given region, then the risk in terms of the chance of seeing an accident on a yearly basis in
that region can be derived. This yearly accident rate is of interest as it affects people’s
perception of their exposure to risk, which in turn can affect their willingness to fly, with
obvious commercial implications. This risk measure may also be expressed as a quantity
analogous to the reliability theory parameter “Mean Time Between Failures” (MTBF):
The Mean (average) Time Between Accidents (MTBA) is just the reciprocal of the
average number of accidents per year. For example, for an average of three accidents per
year, the MTBA is 1/3 year, or about 122 days.
The choice of a risk metric also depends upon the type of risk one is attempting to
measure. For example, for midair collisions en route, an appropriate activity measure
would be flight hours. However, for collisions in the terminal environment, a more
appropriate measure of activity would be the number of departures and arrivals.
SEPARATION SAFETY MODELING
3-4
Table 3-1
Chance of Collision per Year, for Constant Assumed Risk
per Flight Hour, by Region and Year2
Region Australia (Tasman3) North Atlantic
MNPS
ECAC
Approximate number of
en route commercial flight
hours for 1995
30,600 835,500 5,778,000
Assumed level of risk
expressed in fatal
accidents per flight hr
1.5x10-8 1.5x10-8 1.5x10-8
Assumed level of risk
expressed in chance of
collision4 per year, 1995
1
4357
1
160
1
23
Assumed level of risk
expressed in chance of
collision per year, 2000
1
3565
1
116
1
19
Table 3-2
Risk per Flight Hour by Region for Constant Assumed Chance
of Collision per Year, Based on 1995 Traffic Estimates
Collision Risk per Flight Hour
Assumed Chance of
Collision per Year
Australia (Tasman) North Atlantic
MNPS
ECAC
1/4537 7.2x10-9 2.6x10-10 3.8x10-11
1/160 2.0x10-7 7.4x10-9 1.1x10-9
1/23 1.4x10-6 5.2x10-8 7.5x10-9
The user of the risk metric often determines what is the most appropriate metric. For
example, a regulatory, standard-setting authority might consider the number of (fatal)
2 1995 flight hours, and predicted increase for year 2000, obtained as follows:
Australian: Personal communication with Australian CAA
NAT MNPS: NAT Traffic Forecasting Report
ECAC: Eurocontrol Air Traffic Statistics and Forecasts Doc No 94.80.14
3 Roughly speaking, the area between 24°S to 44°S and 174°E to 151°E.
4 It is assumed that one collision equals two fatal accidents.
COLLISION MODELING DEFINITIONS AND RISK METRICS
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
a concept paper for separation safety modeling(12)
