Automated and Intelligent Structure Design Generation and Exploration

Abstract

Computational systems and methods are disclosed that learn about and assist with appropriate structure-based design choices. The systems and methods autonomously explore design states, and suggest to a user one or more optimize design states. Constraints are provided to limit exploration to valid design states. Systems and methods are disclosed that assist groups of users with coordinating their efforts in producing a cohesive design. Systems and methods are disclosed for learning from past optimizations in order to provide more rapid convergence on an optimized design, avoid local maxima and other hurdles to optimization, avoid undesired optimizations, and so on.

What is claimed is: 
     
         1 . A computer-implemented system for developing a structure design, comprising:
 a design engine which receives various inputs, and produces a first structure design state;   an attributes engine, configured to:
 receive a plurality of structure design states including said first structure design state, each said design state comprised of a plurality of cells, each cell configured to permit it to interconnect with another cell such that a structure design state comprised of cells represents a valid structure design state; 
 quantify a plurality of measures of various attributes of each said structure design state; 
 from said quantified plurality of measures of various attributes of each said structure design state, determine a structure fitness function value corresponding to said first structure design state; 
   an optimization engine, communicatively connected to said design engine and said attributes engine, configured to:
 receive, from said design engine, said first structure design state; 
 receive, from a change operation selection mechanism, a first change operation; 
 apply said first change operation to said first structure design state to thereby obtain a second structure design state; 
 provide said second structure design state to said attributes engine to thereby obtain a structure fitness function value corresponding to said second structure design state; and 
 receive and compare said structure fitness function values corresponding to said first and second structure design states and designating as a selected design state that design state having a greater fitness function value and designating as a non-selected design state that design state having a lesser fitness function value. 
   
     
     
         2 . The computer-implemented system of  claim 1 , wherein said design engine, said attributes engine, said optimization engine, and said learning engine are configured to apply change operations to multiple design states, determine a corresponding structure fitness function value for each result of each said application of a change operation to a design state, compare all said fitness function values so as to determine a design state having a maximized fitness function value, further comprising:
 a design workspace configured to present to a user said design state having a maximized fitness function value.   
     
     
         3 . The computer-implemented system of  claim 2 , further comprising an interface for providing an indication to a user of said structure fitness function value for said design state having a maximized fitness function value. 
     
     
         4 . The computer-implemented system of  claim 1 , wherein said design engine is configured such that a user may select said first structure design state. 
     
     
         5 . The computer-implemented system of  claim 4 , wherein said system includes a control permitting said user to designate a user-created design state as said first structure design state. 
     
     
         6 . A computer-implemented system for developing a structure design, comprising:
 a design engine which receives various inputs, and produces a first structure design state;   an attributes engine, configured to:
 receive a plurality of structure design states including said first structure design state; 
 quantify a plurality of measures of various attributes of each said structure design state; 
 from said quantified plurality of measures of various attributes of each said structure design state, determine a corresponding structure fitness function value,
     F=f ( ŵ   1   â   1   ,ŵ   2   â   2   , . . . ŵ   n   â   n ) 
 
   wherein a 1 , a 2 , . . . a n  are each a quantification of an attribute, respectively, of said corresponding structure design state, and w 1 , w 2 , . . . w n  are each a weighting value corresponding to each said attribute quantification, respectively;   an optimization engine, communicatively connected to said design engine and said attributes engine, configured to:
 receive, from said design engine, said first structure design state; 
 receive, from a change operation selection mechanism responsive to preferred change operations, a first change operation; 
 apply said first change operation to said first structure design state to thereby obtain a second structure design state; 
 provide said second structure design state to said attributes engine to thereby obtain a corresponding structure fitness function value; 
 receive and compare said structure fitness function values corresponding to said first and second structure design states and designating as a selected design state that design state having a greater fitness function value and designating as a non-selected design state that design state having a lesser fitness function value; and 
   a learning engine communicatively coupled to said design engine and said optimization engine and configured to update said change operation selection mechanism such that the change operation producing the selected design state is provided with a preference as compared to the change operation producing the nonselected design state.   
     
     
         7 . The computer-implemented system of  claim 6 , wherein said change operation selection mechanism selects said change operation from a change operation memory, said change operation memory being user-modifiable. 
     
     
         8 . The computer-implemented system of  claim 6 , wherein said design engine, said attributes engine, said optimization engine, and said learning engine are configured to apply change operations to multiple design states, determine a corresponding structure fitness function value for each result of each said application of a change operation to a design state, compare all said fitness function values so as to determine a design state having a maximized fitness function value, further comprising:
 a design workspace configured to present to a user said design state having a maximized fitness function value.   
     
     
         9 . The computer-implemented system of  claim 8 , further comprising an interface for providing an indication to a user of said structure fitness function value for said design state having a maximized fitness function value. 
     
     
         10 . The computer-implemented system of  claim 6 , further comprising a validity checking mechanism for determining whether the second structure design state is a valid design state, and if it is not a valid design state:
 said optimization engine:   receiving a second change operation;   applying said second change operation to said first structure design state to thereby obtain a third design state;   providing said third state to said attributes engine to thereby obtain a corresponding structure fitness function value;   receiving and comparing said structure fitness function values corresponding to said first and third design states and designating as a selected design state that design state having a greater fitness function value and designating as a nonselected design state that design state having a lesser fitness function value;   said learning engine updating said change operation selection mechanism such that the change operation producing the selected design state is provided with a preference as compared to the change operation producing the non-selected design state.   
     
     
         11 . The computer-implemented system of  claim 10 , wherein said checking mechanism for determining whether the second structure design state is a valid design state further comprises a validity rules memory, said validity rules memory being user-modifiable. 
     
     
         12 . The computer-implemented system of  claim 10 , wherein said learning engine is configured to update said change operation selection mechanism such that a combination of the change operation and first structure design state producing the invalid second structure design state is provided with a negative preference. 
     
     
         13 . The computer-implemented system of  claim 6 , wherein said design engine is configured to select said first structure design state. 
     
     
         14 . The computer-implemented system of  claim 6 , wherein said design engine is configured such that a user may select said first structure design state. 
     
     
         15 . The computer-implemented system of  claim 14 , wherein said system includes a control permitting said user to designate a user-created design state as said first structure design state. 
     
     
         16 . The computer-implemented system of  claim 6 , wherein said change operation is a compound change operation comprising a plurality of primitive change operations. 
     
     
         17 . The computer-implemented system of  claim 6 , wherein said first change operation is configured so as to avoid an invalid design state. 
     
     
         18 . The computer-implemented system of  claim 6 , wherein said fitness function value is assigned as a lowest fitness function value if said second structure design state is an invalid design state. 
     
     
         19 . A computer-implemented system for developing a structure design, comprising:
 a design engine, comprising;
 a user interface permitting a user to create a first structure design state; 
 a user interface permitting a user to modify said first structure design state to thereby obtain a second structure design state by way of application of a change operation to said first structure design state, said change operation having a first preference value; 
   a learning engine, communicatively coupled to said design engine, configured to increase said first preference value associated with a change operation for each instance that said change operation is applied such that said change operation represents a preferred change operation for future design states.   
     
     
         20 . The computer-implemented system of  claim 19 , wherein an authorized user may manually adjust said first preference value for said change operation.