Method and apparatus for filtering out signal components from output signals of a differential dosing scale using a fuzzy logic filter
AssigneeCarl Schenck AG
InventorsJosef Wehhofer, Hans Werner, Volker Tews
Life of this patent
Estimating the term…
Loading patent…
AssigneeCarl Schenck AG
InventorsJosef Wehhofer, Hans Werner, Volker Tews
Estimating the term…
The accuracy of the output feed rate or output feed quantity of a differential dosing scale is increased by eliminating from the respective feed rate signal or output quantity signal measuring disturbance that can adversely influence the output accuracy. For this purpose, the size of the output signal variations or of the feed rate variations are processed in a fuzzy logic circuit (9) which allocates the respective signals to fuzzy sets, e.g. "small", "medium", "large" in accordance with type and size of the disturbance or interference. The fuzzy logic circuit (9) forms a filter control signal that corresponds to a filtering degree or filter factor. The filtering control signal controls the time constant of a time variable filter (6) through which the differentiated signal coming from the dosing scale load cells is passed prior to processing in the fuzzy logic circuit (9).
A method for filtering out measurement falsifying signal components from measured output signals of a differential dosing scale, comprising the following steps:
An apparatus for filtering out measurement falsifying signal components from actual measured scale output signals of a differential dosing scale, comprising a differential dosing scale including a feed drive and weight sensors (3) for measuring actual scale output signal variations to provide size variations representing values, and a signal evaluating and control circuit (4) for processing said size variations representing values, said control circuit comprising a differentiating circuit (5) having an input connected to said weight sensors through a conductor (3A) for producing an actual feed rate signal at an output (5A, 5B) of said differentiating circuit (5), a time variable filter (6) having a first input (6A) connected to said output (5B) of said differentiating circuit (5), said time variable filter (6) having a first output (6B) providing a future prognosis value and a second output (6C) providing a current estimated value, a first comparator circuit (7) having a first input connected to said output (5A) of said differentiating circuit (5) to receive said actual feed rate signal and a second input connected to said first output (6B) of said time variable filter (6) to receive said future prognosis value for producing a first difference value (at 7A) between said future prognosis value and said actual feed rate signal at a signal sampling time, said signal evaluating and control circuit (4) further comprising a fuzzy logic circuit (9) having a first input (9A) connected to an output (7A) of said first comparator circuit (7) for evaluating said first difference value for its size to provide size variations representing values and for allocating said size variations representing values to different value groups representing the type and size of a disturbance that causes said measurement falsifying signal components, said fuzzy logic circuit producing a filter control signal at its control output (9B) in response to said value groups, said time variable filter (6) having a second control input (6D) connected through a conductor (9F) to said control output (9B) of said fuzzy logic circuit (9) for controlling said time variable filter (6) to filter out said measurement falsifying signal components in response to said filter control signal so that said estimated value at said second output (6C) of said filter (6) is free of said falsifying signal components.
Dependencies are inferred from the claim text.