Tuning model structures of dynamic systems

Abstract

Tuning model structures of dynamic systems are described herein. One method for tuning model structures of a dynamic system includes predicting a variable for each of a number of models associated with a number of model structures of a dynamic system, calculating a rate of error of the predicted variable for each of the number of models compared to an observed variable, determining a best model structure among the number of model structures based on the calculated rate of error, and creating a revised model structure using the best model structure to tune the number of model structures of the dynamic system.

What is claimed: 
     
         1 . A non-transitory computer readable medium having computer readable instructions stored thereon that are executable by a processor to:
 determine, based on a rate of error associated with each of a plurality of models associated with a plurality of model structures, a best model structure and a worst model structure among the plurality of model structures;   generate a revised model structure using the best model structure by combining a sub-group of the plurality of model structures;   replace the worst model structure with the revised model structure; and   control a heating, ventilation, and air-conditioning (HVAC) system using the best model structure.   
     
     
         2 . The computer readable medium of  claim 1 , wherein the computer readable instructions are executable by the processor to define a plurality of new model structures eligible to be the best model structure. 
     
     
         3 . The computer readable medium of  claim 1 , wherein generating the revised model structure includes migrating a random model structure towards the best model structure. 
     
     
         4 . The computer readable medium of  claim 1 , wherein the revised model structure includes a model structure with a calculated rate of error associated therewith that is less than a calculated rate of error associated with the best model structure. 
     
     
         5 . The computer readable medium of  claim 1 , wherein the computer readable instructions are executable by the processor to calculate the rate of error by comparing a plurality of predicted variables to a plurality of observed variables for each of the plurality of models associated with the plurality of model structures. 
     
     
         6 . The computer readable medium of  claim 1 , wherein the computer readable instructions are executable by the processor to receive a prioritization of a plurality of variables from a user. 
     
     
         7 . The computer readable medium of  claim 1 , wherein the computer readable instructions are executable by the processor to update the rate of error associated with each of the plurality of models over time. 
     
     
         8 . The computer readable medium of  claim 1 , wherein generating the revised model structure includes creating a new model structure using surrogate modeling. 
     
     
         9 . A computing device for tuning model structures, comprising:
 a memory; and   a processor configured to execute executable instructions stored in the memory to:
 determine, based on an evaluation of a plurality of model structures associated with a plurality of models, a best model structure and a worst model structure among the plurality of model structures; 
 generate a revised model structure using the best model structure by combining a sub-group of the plurality of model structures; 
 replace the worst model structure with the revised model structure; and 
 control a heating, ventilation, and air-conditioning (HVAC) system by using the best model structure. 
   
     
     
         10 . The computing device of  claim 9 , wherein the instructions are executable by the processor to evaluate the plurality of model structures. 
     
     
         11 . The computing device of  claim 9 , wherein the worst model structure is a model structure with a higher calculated rate of error than the remaining plurality of model structures. 
     
     
         12 . The computing device of  claim 9 , wherein the worst model structure is a model structure with less than a threshold plurality of evaluations. 
     
     
         13 . The computing device of  claim 9 , wherein the plurality of model structures include dependencies between the plurality of variables for each of the plurality of models. 
     
     
         14 . The computing device of  claim 9 , wherein the computing device is included in a supervisory control system associated with controllers of the HVAC system. 
     
     
         15 . A method for tuning model structures, comprising:
 identifying, by a computing device, a plurality of models associated with a plurality of model structures;   evaluating, by the computing device, the plurality of model structures;   identifying, by the computing device, a best model structure and a worst model structure based on the evaluation of the plurality of model structures;   generating, by the computing device, a revised model structure using the best model structure by combining a sub-group of the plurality of model structures;   replacing, by the computing device, the worst model structure with the revised model structure;   tuning, by the computing device, the plurality of model structures by defining a plurality of new model structures eligible to be the best model structure; and   controlling, by the computing device, a heating, ventilation, and air-conditioning (HVAC) system by using the best model structure.   
     
     
         16 . The method of  claim 15 , wherein evaluating the plurality of model structures includes:
 calculating a predicted value for a plurality of variables in a model associated with a model structure over a time period, wherein the plurality of model structures have dependencies between the plurality of variables for each of the plurality of models;   comparing the predicted value for the plurality of variables to an observed value for each of the plurality of variables over the time period; and   calculating a rate of error for the model structure based on the comparison of the predicted values and the observed values.   
     
     
         17 . The method of  claim 15 , wherein the method includes saving a plurality of well performing model structures, wherein the plurality of well performing model structures include a plurality of model structures with a lower calculated rate of error than a remaining plurality of model structures. 
     
     
         18 . The method of  claim 17 , wherein generating the revised model structure includes migrating a well performing model structure among the plurality of well performing model structures towards the best model structure. 
     
     
         19 . The method of  claim 15 , wherein the method includes saving a plurality of immature model structures, wherein the plurality of immature model structures include a plurality of model structures with less than a threshold plurality of evaluations. 
     
     
         20 . The method of  claim 19 , wherein the method includes refraining from replacing the plurality of immature model structures.