초록 ▼

A method of obtaining improved feedforward data for a feedforward control system to move a component through a setpoint profile is presented. The setpoint profile includes a plurality of target states of the component each to be substantially attained at one of a corresponding sequence of target tim

A method of obtaining improved feedforward data for a feedforward control system to move a component through a setpoint profile is presented. The setpoint profile includes a plurality of target states of the component each to be substantially attained at one of a corresponding sequence of target times. The method includes moving the component with the feedforward control system according to the setpoint profile using a first set of feedforward data; measuring a state of the component at a plurality of times during the movement; comparing the measured states with corresponding target states defined by the setpoint profile to obtain a set of errors; filtering the set of errors with a non-linear filter; generating improved feedforward data based on the filtered errors, the improved feedforward data being usable by the feedforward control system to move the component more accurately through the setpoint profile.

대표청구항 ▼

What is claimed is: 1. A method of obtaining improved feedforward data for a feedforward control system to move a component through a setpoint profile, said setpoint profile comprising a plurality of target states of said component each to be substantially attained at one of a corresponding sequenc

What is claimed is: 1. A method of obtaining improved feedforward data for a feedforward control system to move a component through a setpoint profile, said setpoint profile comprising a plurality of target states of said component each to be substantially attained at one of a corresponding sequence of target times, said method comprising: a) moving said component with the feedforward control system according to the setpoint profile using a first set of feedforward data; b) measuring a state of said component at a plurality of times during said moving; c) comparing the measured states with corresponding target states defined by said setpoint profile to obtain a set of errors; d) comparing the set of errors to a threshold and determining if the set of errors or a subset of the set of errors are above the threshold; e) filtering said set of errors with a dead zone non-linear filter if the set of errors or the subset of the set of errors are above the threshold; f) generating improved feedforward data based on the filtered errors, said improved feedforward data being usable by said feedforward control system during a subsequent moving of said component more accurately through the setpoint profile; g) repeating steps (a) to (f) iteratively, using the most recently obtained improved feedforward data instead of the first set of feedforward data, until it is determined that the set of errors or the subset of the set of errors obtained in step (c) are below the threshold. 2. The method according to claim 1, wherein the filtering is dependent on the amplitudes of said set of errors. 3. The method according to claim 2, wherein the filtering includes suppressing errors of small amplitude more than errors of larger amplitude. 4. The method according to claim 1, wherein the filtering includes decomposing the error signal into a plurality of discrete frequency bands and individually applying a selected filtering operation to each of the bands. 5. The method according to claim 1, wherein the plurality of times during the movement at which the state of the component is measured correspond to at least a subset of said plurality of target times. 6. The method according to claim 1, wherein said target states include at least one of the following: target positions of said component, target velocities of said component, and target accelerations of said component. 7. The method according to claim 1, wherein said filter is configured to pass contributions to the errors that have a magnitude greater than a predefined noise bound and to reject other contributions. 8. The method according to claim 1, wherein if the set of errors or the subset of the set of errors are below the predetermined threshold, the first set of feedforward data is used as improved feedforward data during a subsequent moving of said component. 9. A method of obtaining improved feedforward data for a feedforward control system to move a component through a setpoint profile, said setpoint profile comprising a plurality of target states of said component each to be substantially attained at one of a corresponding sequence of target times, said method comprising: a) moving said component with the feedforward control system according to the setpoint profile using a first set of feedforward data; b) measuring a state of said component at a plurality of times during said moving; c) comparing the measured states with corresponding target states defined by said setpoint profile to obtain a set of errors; d) filtering said set of errors with a non-linear filter; and e) generating improved feedforward data based on the filtered errors, said improved feedforward data being usable by said feedforward control system during a subsequent moving of said component more accurately through the setpoint profile, wherein said set of errors is represented by an array of data points ey(i), where 0≦i≦n, n is the number of samples in the set, said filter is represented by an amplitude-dependent filter matrix Φ(ey(i)), and step (d) corresponds to a multiplication of Φ(ey(i)) by ey(i). 10. The method according to claim 9, wherein Φ(ey(i)) is defined as follows: Φ ⁢ ⁢ ( e _ y ) = [ ϕ ( e _ y ⁡ ( 1 ) ) 0 ⋯ 0 0 ϕ ( e _ y ⁡ ( 2 ) ) ⋯ 0 0 ⋮ ⋱ ⋮ 0 0 ⋯ ϕ ( e _ y ⁡ ( n ) ) ] with ϕ ⁡ ( x ) = ɛ ⁡ ( x ) - δɛ ⁡ ( x )  x  and ɛ ⁡ ( x ) = { 1 , if ⁢ ⁢  x  ≥ δ 0 , if ⁢ ⁢  x  < δ . 11. A lithographic projection apparatus arranged to project a pattern from a patterning device onto a substrate, comprising: (a) a movable support configured to support a component; and (b) a system configured to move said movable support through a setpoint profile comprising a plurality of target states of said movable support, each to be substantially attained at one of a corresponding sequence of target times, said system comprising: (i) a displacement device configured to move said movable support according to said setpoint profile; (ii) a feedforward control system configured to control said displacement device using a first set of feedforward data; (iii) a measuring system configured to measure the state of said component at a plurality of times during movement of said movable support; (iv) a comparison device configured to compare the measured states with corresponding target states to obtain a set of errors, said set of errors being compared to a threshold to determine if the set of errors or a subset of the set of errors are above the threshold; (v) a dead zone non-linear filter configured to filter said set of errors if the set of errors or a subset of the set of errors are above the threshold; (vi) a feedforward data generating device arranged to generate modified feedforward data based on the filtered set of errors, said modified feedforward data being usable by said feedforward control system during a subsequent moving of the movable support to control said displacement device to more accurately move said movable support through the setpoint profile, said system being configured to iteratively use the most recently obtained modified feedforward data instead of the first set of feedforward data to determine improved feedforward data until it is determined that the set of errors or the subset of the set of errors obtained by the comparison device are below the threshold. 12. The lithographic apparatus according to claim 11, wherein said component is the substrate or the patterning device. 13. A device manufacturing method, comprising: (a) projecting a pattern from a patterning device onto a substrate; (b) supporting a component with a movable support; (c) moving the movable support with a feedforward control system through a setpoint profile using a first set of feedforward data, said setpoint profile comprising a plurality of target states of said movable support, each to be substantially attained at one of a corresponding sequence of target times; (d) measuring a state of the component at a plurality of times during said moving; (e) comparing the measured states with corresponding target states to obtain a set of errors; (f) comparing the set of errors to a threshold and determining if the set of errors or a subset of the set of errors are above the threshold; (g) filtering the set of errors with a dead zone non-linear filter if the set of errors or a subset of the set of errors are above the threshold; (h) generating improved feedforward data based on the filtered set of errors; (i) moving the movable support through the setpoint profile using the improved feedforward data; (j) repeating steps (c) to (h) iteratively, using the most recently obtained improved feedforward data instead of the first set of feedforward data, until it is determined that the set of errors or subset of the set of errors obtained in step (e) are below the threshold. 14. The method according to claim 13, wherein the component is the substrate or the patterning device. 15. The method according to claim 13, wherein said set of errors is represented by an array of data points ey(i), where 0≦i≦n, n is the number of samples in the set, said filter is represented by an amplitude-dependent filter matrix Φ(ey(i)), and step (d) corresponds to a multiplication of Φ(ey(i)) by ey(i). 16. The lithographic apparatus according to claim 11, wherein said set of errors is represented by an array of data points ey(i), where 0≦i≦n, n is the number of samples in the set, said filter is represented by an amplitude-dependent filter matrix Φ(ey(i)), and step (d) corresponds to a multiplication of Φ(ey(i)) by ey(i).