Abstract
The performance of non-linear identification techniques is often determined by the appropriateness of the selected input variables and the corresponding time lags. High correlation coefficients between candidate input variables in addition to a non-linear relation with the output signal induce the need for an appropriate input selection methodology. This paper proposes a genetic polynomial regression technique to select the significant input variables for the identification of non-linear dynamic systems with multiple inputs. Statistical tools are presented to visualize and to process the results from different selection runs. The evolutionary approach can be used for a wide range of identification techniques and only requires a minimal input and a priori knowledge from the user. The evolutionary selection algorithm has been applied on a real-world example to illustrate its performance. The engine load in a combine harvester is highly variable in time and should be kept below an allowable limit during automatic ground speed control mode. The genetic regression process has been used to select those measurement variables that have a significant impact on the engine load and that will act as measurement variables of a non-linear model-based engine load controller.
Similar content being viewed by others
Abbreviations
- Symbol:
-
Description
- A:
-
Matrix of p+1 polynomial model parameters
- a i :
-
ith polynomial model parameter
- C p :
-
Mallow’s Cp
- d :
-
Polynomial degree
- LI(x i ):
-
Lifetime Index (%)
- n :
-
number of candidate input variables
- N min :
-
Final number of regressor variables
- N runs :
-
Number of selection runs
- n tot :
-
Initial number of candidate regressor variables
- N :
-
Number of data samples
- R2 p :
-
Adjusted coefficient of determination
- RMSE p :
-
Residual Mean Squared Error
- R uu (τ):
-
Auto-correlation function
- R yu (τ):
-
Cross-correlation function
- sd j :
-
Combination of maximum d input variables
- X :
-
Initial set of n tot candidate input variables
- x I :
-
Candidate input variable i
- X sel (n):
-
Selected set of n input variables
- y :
-
Output signal
- γ yu :
-
Normalized cross-correlation coefficient
- ρ uu (τ):
-
Auto-correlation coefficient
- σ 2 yy , σ 2 uu :
-
Variance of output signal y and input signal u
- τ:
-
Time lag (s)
References
Chiu S L (1996) Selecting input variables for fuzzy models. J Intell Fuzzy Syst 3:271–279
Gaweda AE, Zurada, JM, Setiono R (2001) Input selection in data-driven fuzzy modeling. In: Proceedings of the 10th IEEE international conference of fuzzy systems, Melbourne, Australia
Lin Y, Cunningham III GA, Coggeshall SV (1996) Input variable identification – fuzzy curves and fuzzy surfaces. Fuzzy Sets Syst 82:65–71
Jang J-SR (1996) Input selection for ANFIS learning. In: Proceedings of the 5th IEEE international conference on fuzzy systems, New Orleans, USA
Koza JR (1992) Genetic programming – on the programming of computers by means of natural selection. MIT, Cambridge, MA, USA
Maertens K, De Baerdemaeker J, Ramon H, De Keyser R (2001) An analytical grain flow model of a combine harvester. part 1: design of the model. J Agric Eng Res 79:55–63
Maertens K, Johansen TA, Babuška R (2004) Engine load prediction in off-road vehicles using multi-objective nonlinear identification. Control Engineering Practice (In Press).
Maertens K, Ramon H, De Baerdemaeker J (2003) An on-the-go modelling algorithm for separation processes in combine harvesters. Comput Electron Agric 43(3):197–207
Mallows CL (1973) Come comments on Cp. Technometrics 15:661–675
Neter J, Kutner MH, Nachtstheim CJ, Wasserman W (1996) Applied linear statistical models, 4th edn. McGraw-Hill, New York, USA
Sugeno M, Yasukawa (1993) A fuzzy-logic-based approach to qualitative modeling. IEEE Trans Fuzzy Syst 1:7–31
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Maertens, K., Baerdemaeker, J.D. & Babuška, R. Genetic polynomial regression as input selection algorithm for non-linear identification. Soft Comput 10, 785–795 (2006). https://doi.org/10.1007/s00500-005-0008-8
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-005-0008-8