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
Registration Board (ARB), the U.K. certifying authority, in the late 1950s, when the issue
of approving an automatic “all-weather” landing system for passenger aircraft arose.
James Doolittle had made “blind” landings in 1927, and the French Aero Postale had
SEPARATION SAFETY MODELING
2-4
safely made blind landings for years, but such capability had not been approved by any
country for airliners.
The ARB started by looking at history, establishing the landing accident rate then being
achieved. Based on a study of nearly twenty airline landing accidents (mostly propeller
aircraft), it found that there had been about one accident for each million landings. It
reasoned that introduction of a new capability such as automatic landing would be
designed with the intent of improving the safety record, but could, in no event, allow
safety to deteriorate. It called for a design that would have a predicted (and, to the extent
possible, demonstrated) failure rate of no more than one in ten million landings, ten times
better that the rate experienced in normal operations. It imposed the assumption that any
failure of the automatic landing system would result in a fatal accident - a pessimistic
assertion.
The ARB was aware of another basic point: The kind of analysis it was imposing had
little or no value in offering guidance on the absolute safety to be achieved, but it was
valuable in comparing the safety value of several alternative approaches. The idea of a
target level of safety was attractive and felt philosophically right to many in the industry.
It was rooted in the reality of safety actually achieved, based on historical record.
Building a realistic goal for improvement was easy and could be based on rational
grounds. Application of the target level of safety poses tough problems, not the least of
which is the collection of adequate data, but its implication - that any new design, whether
it is an engine, flight control system, or wing structure, must be at least as good and
hopefully better, than its predecessors - feels right.
Many people and organizations tested the target level of safety idea, especially the target
number of one accident in ten million events. A study done by the International Civil
Aviation Organization Review of the General Concept of Separation Panel (ICAO
RGCSP) in 1975, using UK mortality rates, showed that the risk of mortality in the
healthiest age groups was six in ten million person hours. By comparison, 1965-1973 data
yielded a value of 10.5 fatal accidents in 10 million flying hours.
ICAO, the U.N. international aviation standard-setting body, published further
corroborating information. A comprehensive study by several countries, evaluated fatality
rates in manufacturing, railway work, and public road vehicles; mortality rates in the
general populations; and a variety of air accidents, from landings to midair collisions. The
resulting finding was that an appropriate target level of safety might be between one and
six fatal accidents in 10 million flying hours, with the resulting risk appropriately shared
among mechanical failure, midair collisions, and other accident causes. This target range
seemed credible: It should be used with caution, but it was rational.
ESTABLISHING SEPARATION STANDARDS
2-5
2.1.3 Recent Events
Important strides have been made by the FAA in recent years in reducing longitudinal and
lateral separations. The case was well described in 1988 by the Airport Associations
Coordinating Council (AACC): “...Existing separation standards are highly conservative,
as they should be and must be. While they must always be conservative, current standards
make little or no distinction between situations where highly sophisticated aircraft/air
crew/air traffic service/equipment infrastructures exist, and situations where only basic
services may be available. The current separation figures may have been agreed upon, in
part to protect against difficulties at airports where the quality of air traffic control may
not be as good as might be desirable. It seems to us that a way must be found to provide
the highest level of safety protection for all circumstances without putting at a
disadvantage airports that have adequate air traffic control service, and which badly need
all the capacity that can safely be achieved.”
There are large contrasts between separations allowed in practice in different countries
and at different airports for such procedures as independent simultaneous instrument flight
rules (IFR) approaches to parallel runways, dependent parallel IFR approaches, required
IFR longitudinal separations, and converging IFR approaches. Perhaps the most
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(8)
