Page 72 - Methods For Monitoring And Diagnosing The Efficiency Of Catalytic Converters A Patent - oriented Survey

54 Methods for Monitoring and Diagnosing the Eficiency of Catalytic Converters

In DE4139560 (1993) a method is presented, which is applicable at stationary conditions of
the engine (e.g. idling). A number of n oscillations of the upstream 11 signal takes place
between a rich and a lean value. The product of the two 1 signals is formed

P = A,
.
Then the sum of all these products is formed and the average value is calculated

If AZ<E, where E is a predefined threshold, then the catalytic converter is considered to be of
low efficiency.

For a higher precision, the upstream signal can be correlated with a phase shift with the
downstream signal. This correlation value is used for evaluating the performance loss of the
catalytic converter. This can be better explained by means of fig. 23, where the airhe1 ratio
variation measured by the upstream sensor (fig. 23a) and the corresponding airhel ratio
variation measured by the downstream sensor are shown for the case of a slightly deteriorated
catalytic converter (fig. 23b) and for the case of a greatly deteriorated catalytic converter (fig.
23c). A phase shift r is present between hl and l.2, which is essentially dependent upon the
running time of the exhaust gas from the upstream sensor to the downstream sensor. This
phase shift r is hardly dependent upon converter performance loss. The simplest correlation
used is the cross correlation by multiplying the upstream phase shifted signals hl(t + t) with the
downstream measured signals b(t) and by summing the individual products.

Another correlation used is the orthogonal correlation. The signals Al(t) and hz(t) are
multiplied with each other in order to obtain a real component RE after averaging. The
upstream signal is then shifted by a quarter period (p/4) with respect to the output signal and
the two signals are multiplied and an imaginary component is obtained. The amplitude and the
phase p of the output signal can then be computed from the real and imaginary components.
All four signals, that is the real and imaginary components, the amplitude and the phase p can
each be used individually for evaluating the state of the performance loss of the catalytic
converter.

In patent application Ern546318 (1993) the following method is considered. At time instant to
it is assumed that the oxygen deficient input quantity in the catalytic converter is higher than
the oxygen stored. The quantity of fuel injected in the engine then starts oscillating. The engine
air/f%el modulation factor I*R has a square wave form and the mean value slowly runs down in
time towards a lean shift of mean aidflow ratio (fig. 24a). The hatched areas over 1 indicate a
deficiency of oxygen, while the areas under /.'I(= 1 represent an excess oxygen input quantity.