Parameter tuning

Abstract

A method, performed by a computer device, may include interacting with one or more plots associated with a simulation of an executable graphical model, receiving a selection of one or more target output points associated with the one or more plots; receiving a selection identifying one or more parameters, associated with the one or more plots, of one or more system parameters, associated with the executable graphical model; and generating an inverse parameter Jacobian matrix. The inverse parameter Jacobian matrix may correspond to an inverse of a parameter Jacobian matrix that relates one or more partial derivatives of the one or more system parameters to one or more system equations. The method may further include determining one or more tuning values for the selected one or more parameters based on the generated inverse parameter Jacobian matrix and based on the selected one or more target output points.

1 . A method, performed by a computer device, the method comprising:
 interacting with one or more plots associated with a simulation of an executable graphical model, wherein the interacting is performed by the computer device;   receiving a selection of one or more target output points associated with the one or more plots, wherein the receiving is performed by the computer device;   receiving a selection identifying one or more parameters, associated with the one or more plots, of one or more system parameters, associated with the executable graphical model, wherein the receiving is performed by the computer device;   generating an inverse parameter Jacobian matrix, wherein:
 the inverse parameter Jacobian matrix corresponds to an inverse of a parameter Jacobian matrix associated with the executable graphical model, 
 the parameter Jacobian matrix relates one or more partial derivatives of the one or more system parameters, associated with the executable graphical model, to one or more system equations, associated with the executable graphical model, and 
 the generating is performed by the computer device; and 
   determining one or more tuning values for the selected one or more parameters based on the generated inverse parameter Jacobian matrix and based on the selected one or more target output points, wherein the determining is performed by the computer device.   
     
     
         2 . The method of  claim 1 , wherein determining the inverse parameter Jacobian matrix includes:
 determining the one or more system parameters associated with the executable graphical model;   determining the one or more partial derivatives of the one or more system parameters with respect to the one or more system equations;   determining the parameter Jacobian matrix based on the determined one or more partial derivatives; and   inverting the parameter Jacobian matrix to generate the inverse parameter Jacobian matrix.   
     
     
         3 . The method of  claim 1 , further comprising:
 determining that the number of the selected one or more target output points equals the number of the selected one or more parameters;   determining that a unique solution exists to the one or more tuning values for the selected one or more parameters, based on determining that the number of the selected one or more target output points equals the number of the selected one or more parameters; and   providing the one or more tuning values for the selected one or more parameters as the determined unique solution.   
     
     
         4 . The method of  claim 1 , further comprising:
 determining that the number of the selected one or more target output points is less than the number of the selected one or more parameters; and   determining that more than one solutions exists to the one or more tuning values for the selected one or more parameters, based on determining that the number of the selected one or more target output points is less than the number of the selected one or more parameters.   
     
     
         5 . The method of  claim 4 , further comprising:
 presenting at least a subset of the more than one solutions to the user; and   receiving a selection of one or more solutions, from the at least a subset of the more than one solutions, from the user.   
     
     
         6 . The method of  claim 4 , further comprising:
 selecting one of the more than one solutions based on one or more optimization criteria.   
     
     
         7 . The method of  claim 6 , wherein the one or more optimization criteria include:
 one or more cost functions associated with the executable graphical model; or   a criterion specified by a user in connection with a request to determine the one or more tuning values for the selected one or more parameters.   
     
     
         8 . The method of  claim 6 , wherein selecting the one of the more than one solutions based on the one or more optimization criteria includes:
 determining sensitivities associated with particular ones of the selected one or more parameters;   ranking the selected one or more parameters based on the determined sensitivities;   selecting top n parameters from the ranked selected one or more parameters, wherein n corresponds to the number of the selected one or more target output points; and   generating the inverse parameter Jacobian matrix based on the selected top n parameters.   
     
     
         9 . The method of  claim 8 , wherein a sensitivity associated with a particular one of the selected one or more parameters is based on a value of a partial derivative of the particular one of the selected one or more parameters with respect to the one or more system equations at a time value associated with the selected one or more target output points. 
     
     
         10 . The method of  claim 1 , further comprising:
 determining that the number of the selected one or more target output points is greater than the number of the selected one or more parameters;   determining that no solution exists to the one or more tuning values for the selected one or more parameters, based on determining that the number of the selected one or more target output points is greater than the number of the selected one or more parameters;   computing an approximation to the one or more plots, when no solution exists to the one or more tuning values for the selected one or more parameters, wherein the approximation includes a number of target points equal to the number of the selected one or more parameters; and   determining the one or more tuning values for the selected one or more parameters based on the generated inverse parameter Jacobian matrix and based on the approximation.   
     
     
         11 . The method of  claim 10 , wherein a first one of the one or more target points is associated with a high importance rating and a second one of the one or more target points is associated with a low importance rating, and wherein computing an approximation to the one or more plots includes:
 disregarding the second one of the one or more target points.   
     
     
         12 . The method of  claim 1 , wherein receiving a selection of one or more target responses associated with the one or more plots includes:
 detecting a selection of a point on one of the one or more plots, wherein the point is located within a particular distance from a plot line and within a tolerance limit associated with the plot line.   
     
     
         13 . The method of  claim 1 , wherein receiving a selection of one or more target responses associated with the one or more plots includes:
 receiving a selection to adjust at least one of an amplitude or a frequency associated with one of the one or more plots.   
     
     
         14 . The method of  claim 1 , further comprising:
 providing an indication of one of the determined one or more tuning values, associated with a particular parameter of the selected one or more parameters, in connection with a block associated with the particular parameter.   
     
     
         15 . The method of  claim 1 , further comprising:
 providing an option to generate a source signal based on the inverse parameter Jacobian matrix;   receiving a selection of the option; and   generating the source signal in the executable graphical model, when the selection of the option is received.   
     
     
         16 . The method of  claim 15 , further comprising:
 providing, via the generated source signal, one or more values associated with the inverse parameter Jacobian matrix at a particular time step during a simulation of the executable graphical model.   
     
     
         17 . A non-transitory computer-readable medium containing one or more instructions executable by at least one processor, the computer-readable medium comprising:
 one or more instructions to interact with one or more plots associated with a simulation of an executable graphical model;   one or more instructions to receive a selection of one or more target output points associated with the one or more plots;   one or more instructions to receive a selection identifying one or more parameters, associated with the one or more plots, of one or more system parameters, associated with the executable graphical model;   one or more instructions to generate an inverse parameter Jacobian matrix, wherein:
 the inverse parameter Jacobian matrix corresponds to an inverse of a parameter Jacobian matrix associated with the executable graphical model, and 
 the parameter Jacobian matrix relates one or more partial derivatives of the one or more system parameters, associated with the executable graphical model, to one or more system equations, associated with the executable graphical model; and 
   one or more instructions to determine one or more tuning values for the selected one or more parameters based on the generated inverse parameter Jacobian matrix and based on the selected one or more target output points.   
     
     
         18 . The non-transitory computer-readable medium of  claim 17 , wherein the one or more instructions to determine the inverse parameter Jacobian matrix include:
 one or more instructions to determine the one or more system parameters associated with the executable graphical model;   one or more instructions to determine the one or more partial derivatives of the one or more system parameters with respect to the one or more system equations;   one or more instructions to determine the parameter Jacobian matrix based on the determined one or more partial derivatives; and   one or more instructions to invert the parameter Jacobian matrix to generate the inverse parameter Jacobian matrix.   
     
     
         19 . The non-transitory computer-readable medium of  claim 17 , further comprising:
 one or more instructions to determine that the number of the selected one or more target output points equals the number of the one or more system parameters;   one or more instructions to determine that a unique solution exists to the one or more tuning values for the selected one or more parameters, based on determining that the number of the selected one or more target output points equals the number of the one or more system parameters; and   one or more instructions to provide the one or more tuning values for the selected one or more parameters as the determined unique solution.   
     
     
         20 . The non-transitory computer-readable medium of  claim 17 , further comprising:
 one or more instructions to determine that the number of the selected one or more target output points is less than the number of the one or more system parameters; and   one or more instructions to determine that more than one solutions exists to the one or more tuning values for the selected one or more parameters, based on determining that the number of the selected one or more target output points is less than the number of the one or more system parameters.   
     
     
         21 . The non-transitory computer-readable medium of  claim 20 , further comprising:
 one or more instructions to present at least a subset of the more than one solutions to the user; and   one or more instructions to receive a selection of one or more solutions, from the at least a subset of the more than one solutions, from the user.   
     
     
         22 . The non-transitory computer-readable medium of  claim 20 , further comprising:
 one or more instructions to select one of the more than one solutions based on one or more optimization criteria.   
     
     
         23 . The non-transitory computer-readable medium of  claim 22 , wherein the one or more optimization criteria include:
 one or more cost functions associated with the executable graphical model; or   a criterion specified by a user in connection with a request to determine the one or more tuning values for the selected one or more parameters.   
     
     
         24 . The non-transitory computer-readable medium  22 , wherein the one or more instructions to select the one of the more than one solutions based on the one or more optimization criteria include:
 one or more instructions to determine sensitivities associated with particular ones of the selected one or more parameters;   one or more instructions to rank the selected one or more parameters based on the determined sensitivities;   one or more instructions to select top n parameters from the ranked selected one or more parameters, wherein n corresponds to the number of the selected one or more target output points; and   one or more instructions to generate the inverse parameter Jacobian matrix based on the selected top n parameters.   
     
     
         25 . The non-transitory computer-readable medium  24 , wherein a sensitivity associated with a particular one of the selected one or more parameters is based on a value of a partial derivative of the particular one of the selected one or more parameters with respect to the one or more system equations at a time value associated with the selected one or more target output points. 
     
     
         26 . The non-transitory computer-readable medium of  claim 17 , further comprising:
 one or more instructions to determine that the number of the selected one or more target output points is greater than the number of the selected one or more parameters;   one or more instructions to determine that no solution exists to the one or more tuning values for the selected one or more parameters, based on determining that the number of the selected one or more target output points is greater than the number of the selected one or more parameters;   one or more instructions to compute an approximation to the one or more plots, when no solution exists to the one or more tuning values for the selected one or more parameters, wherein the approximation includes a number of target points equal to the number of the selected one or more parameters; and   one or more instructions to determine the one or more tuning values for the selected one or more parameters based on the generated inverse parameter Jacobian matrix and based on the approximation.   
     
     
         27 . The non-transitory computer-readable medium of  claim 17 , wherein a first one of the one or more target points is associated with a high importance rating and a second one of the one or more target points is associated with a low importance rating, and wherein the one or more instructions to compute an approximation to the one or more plots include:
 one or more instructions to disregard the second one of the one or more target points.   
     
     
         28 . The non-transitory computer-readable medium of  claim 17 , wherein the one or more instructions to receive a selection of one or more target responses associated with the one or more plots include:
 one or more instructions to detect a selection of a point on one of the one or more plots, wherein the point is located within a particular distance from a plot line and within a tolerance limit associated with the plot line.   
     
     
         29 . The non-transitory computer-readable medium of  claim 17 , wherein the one or more instructions to receive a selection of one or more target responses associated with the one or more plots include:
 one or more instructions to receive a selection to adjust at least one of an amplitude or a frequency associated with one of the one or more plots.   
     
     
         30 . The non-transitory computer-readable medium of  claim 17 , further comprising:
 one or more instructions to provide an indication of one of the determined one or more tuning values, associated with a particular parameter of the selected one or more parameters, in connection with a block associated with the particular parameter.   
     
     
         31 . The non-transitory computer-readable medium of  claim 17 , further comprising:
 one or more instructions to provide an option to generate a source signal based on the inverse parameter Jacobian matrix;   one or more instructions to receive a selection of the option; and   one or more instructions to generate the source signal in the executable graphical model, when the selection of the option is received.   
     
     
         32 . (canceled) 
     
     
         33 . The non-transitory computer-readable medium of  claim 31 , further comprising:
 one or more instructions to provide, via the generated source signal, one or more values associated with the inverse parameter Jacobian matrix at a particular time step during a simulation of the executable graphical model.   
     
     
         34 . A computing device, comprising:
 a memory storing executable instructions for implementing an executable graphical model of a dynamic system; and   a processor to:
 interact with one or more plots associated with a simulation of an executable graphical model; 
 receive a selection of one or more target output points associated with the one or more plots; 
 receive a selection identifying one or more parameters, associated with the one or more plots, of one or more system parameters, associated with the executable graphical model; 
 generate an inverse parameter Jacobian matrix, wherein:
 the inverse parameter Jacobian matrix corresponds to an inverse of a parameter Jacobian matrix associated with the executable graphical model, and 
 the parameter Jacobian matrix relates one or more partial derivatives of the one or more system parameters, associated with the executable graphical model, to one or more system equations, associated with the executable graphical model; and 
 
 determine one or more tuning values for the selected one or more parameters based on the generated inverse parameter Jacobian matrix and based on the selected one or more target output points.