Performance data processing system

Claim 1

A method of processing data relating to the performance of apparatus consisting of a series of individual components in order to determine changes in component performance and/or sensor bias from a series of performance measurements from the apparatus by use of a Kalman Filter to analyse changes in said performance measurements, characterised in that a plurality of runs of the Kalman Filter are performed, each succeeding run after the first having the causes of the changes in the performance measurements allocated to a smaller number of component performance changes and;or sensor biasses than the preceding run. the component performance changes and,or sensor biasses to be disregarded for the succeeding run being selected by evaluating a measure of significance for the changes and biasses indicated in the preceding run in order to discard one or more of the least significant changes and or biasses. and selecting as a best answer for the component performance changes and,or sensor biasses. the result of that run in which there is a minimum number of said changes and or biasses that are able to account for the observed performance measurement changes.

Claim 2

A method of processing data relating to the performance of apparatus consisting of a series of individual components in order to determine changes in component performance and/or sensor bias from a series of performance measurements from the apparatus by use of a Kalman Filter to analyse changes in said performance measurements, characterised in that a plurality of runs of the Kalman Filter are performed, each succeeding run after the first having the causes of the changes in the performance measurements allocated to a smaller number of component performance changes and;or sensor biasses than the preceding run. the component performance changes and,or sensor biasses to be disregarded for the succeeding run being selected by evaluating a measure of significance for the changes and biasses indicated in the preceding run in order to discard one or more of the least significant changes and or biasses. and selecting as a best answer for the component performance changes and,or sensor biasses. the result of that run in which there is a minimum number of said changes and or biasses that are able to account for the observed performance measurement changes.

Claim 3

A method of processing data relating to the performance of an apparatus consisting of a series of individual components. in order to determine the changes in component performances and the magnitude of biasses in the measurement system from a series of measurements, comprising the steps of a) defining the a-priori information required by the Kalman Filter, namely: i) a covariance matrix of component performance changes and sensor biasses, ii) a measurement non-repeatability covariance matrix, iii) a System Matrix giving the relationships between observed measurement changes and sensor biasses. b) obtaining performance measurements from the apparatus, c) obtaining datum levels of the performance measurements from the apparatus. d) using the data of steps (b) and (c) to calculate performance measurement differences from datum levels, e) using the data of steps (a) and (d) to derive a most likely set of component changes and sensor biasses by employing a Kalman Filter, the method being characterised by the further steps of f) evaluating an objective function giving a measure of the amount of unexplained measurement change and,or the amount of component change and sensor bias in the solution from (e), g) discarding one or more of the component changes and/or sensor biasses, to leave one or more of said component changes and/or sensor biasses for evaluation, h) repeating steps (e) to (g), if required, until the total number of component changes and sensor biasses left is one, i) considering the sets of estimated component changes and sensor biasses produced in each pass of step (e), together with the corresponding objective function evaluated according to step (f), to identify as a best answer that pass which has produced a minimum number of component changes and sensor biasses that satisfactorily accounts for the observed performance measurement differences.

Claim 4

A method of processing data relating to the performance of an apparatus consisting of a series of individual components. in order to determine the changes in component performances and the magnitude of biasses in the measurement system from a series of measurements, comprising the steps of a) defining the a-priori information required by the Kalman Filter, namely: i) a covariance matrix of component performance changes and sensor biasses, ii) a measurement non-repeatability covariance matrix, iii) a System Matrix giving the relationships between observed measurement changes and sensor biasses. b) obtaining performance measurements from the apparatus, c) obtaining datum levels of the performance measurements from the apparatus. d) using the data of steps (b) and (c) to calculate performance measurement differences from datum levels, e) using the data of steps (a) and (d) to derive a most likely set of component changes and sensor biasses by employing a Kalman Filter, the method being characterised by the further steps of f) evaluating an objective function giving a measure of the amount of unexplained measurement change and,or the amount of component change and sensor bias in the solution from (e), g) discarding one or more of the component changes and/or sensor biasses, to leave one or more of said component changes and/or sensor biasses for evaluation, h) repeating steps (e) to (g), if required, until the total number of component changes and sensor biasses left is one, i) considering the sets of estimated component changes and sensor biasses produced in each pass of step (e), together with the corresponding objective function evaluated according to step (f), to identify as a best answer that pass which has produced a minimum number of component changes and sensor biasses that satisfactorily accounts for the observed performance measurement differences.

Claim 7

A method according to any one of claims 2 to 4 characterised in that, in the best answer selected from step (i). elements whose values are less than a predetermined lower level are discarded.

Claim 8

A method according to any one of claims 2 to 5 characterised in that the average absolute amount of normalised change is evaluated at the end of the first pass and provides a datum for discarding component changes and. or sensor biasses prior to the second pass.

Claim 9

A method according to any one of claims 2 to 6 characterised in that, at the end of the second and each further pass, the changes in the absolute amounts of normalized change relative to the preceding pass are evaluated for each of the remaining component changes and,or sensor biasses and provide datums for discarding component changes and / Or sensor biasses prior to the next pass.

Claim 11

A method according to any one of claims 2 to 8 characterised in that a-priori values for the elements of the covariance matrix of the component performance changes and sensor biasses are estimated and a product is formed with (i) the Kalman Gain Matrix of the Kalman Filter and (ii) the System Matrix, to evaluate a measure of the confusion of sought-for changes or biasses with other changes or biasses due to said estimates, and said estimates are adjusted to reduce the confusion indicated by said evaluation.

Claim 12

A method according to any one of claims 2 to 9 characterised in that the a-priori information supplied to the Kalman Filter includes a best estimate of at least one anticipated component change and or sensor bias.

Claim 13

A method according to any one of the preceding claims characterised in that a plurality of sets of performance measurements are obtained at intervals and are employed to determine levels and/or trends and or vanations from predicted values in component change and/or sensor bias.

Claim 16

An apparatus characterised in that it is provided with means for analysing its performance in accordance with the method of any one of the preceding claims, said means comprising a plurality of performance measurement sensors and evaluation means for processing the data from said sensors by means of said method.