A maximum likelihood rectification (MLR) technique that poses the data
-rectification problem in a probabilistic framework and maximizes the
probability of the estimated plant states given the measurements is pr
oposed. This approach does not divide the sensors into ''normal'' and
''gross error'' classes, but uses all of the data in the rectification
, each sensor being appropriately weighted according to the laws of pr
obability. In this manner, the conventional assumption of no sensor bi
as is avoided and both random errors (noise) and systematic errors (gr
oss errors) are removed simultaneously. A novel technique is introduce
d that utilizes historical plant data to determine a prior probability
distribution of the plant states. This type of historical plant infor
mation, which contains the physical relationships among the variables
(mass balances, energy balances, thermodynamic constraints), as well a
s statistical correlations among the variables, has been ignored in pr
ior data-rectification schemes. This approach can use the historical p
lant information to solve a new class of data-rectification problems i
n which there are no known model constraints. The MLR method is demons
trated on data from a simulated flow network and a simulated heat-exch
anger network. The MLR technique provides considerably improved perfor
mance over existing data-reconciliation schemes in these examples.