KSCE Journal of Civil Engineering

, Volume 23, Issue 12, pp 5194–5200

# Application of Variable Projection Method based on Gram-Schmidt Orthogonalization in Spatial Cartesian Coordinate Transformation Model

• Luyao Wang
• Guolin Liu
• Qiuxiang Tao
• Ke Wang
Surveying and Geo-Spatial Engineering

## Abstract

For the linear and nonlinear parameters that can be separated in the spatial Cartesian coordinate transformation model, we use the variable projection algorithm in this paper to represent the linear parameters with nonlinear parameters, which are transformed into least squares problems with only nonlinear parameters. We simplify the matrix of the nonlinear function by the Gram-Schmidt orthogonalization method, and combine the nonlinear least squares iterative method with the Levenberg-Marquardt (LM) algorithm to solve for the coordinate transformation parameters. Experiments are carried out by solving for the coordinate transformation parameters of the independent spatial Cartesian coordinate system and the CGCS2000 coordinate system. We compare the solution results of the four methods (parameter non-separation method, traditional variable projection method, variable projection method based on QR decomposition, and variable projection method based on Gram-Schmidt orthogonal decomposition) with respect to the calculated results, the number of iterations and the computation time. The experimental results show that the proposed method in this paper requires a lower computation time and achieves higher computational efficiency when obtaining the same solution results and with the same number of iterations.

## Keywords

separable nonlinear least squares problem variable projection gram-schmidt orthogonalization qr decomposition spatial cartesian coordinate transformation

## References

1. Cabrita, C. L., Ruano, A. E., Ferreira, P. M., and Kóczy, L. T. (2013). “Exploiting the functional training approach in Takagi-Sugeno neuro-fuzzy systems.” Soft Comput. Appl. Adv. Intell. Syst. Comput., Vol. 195, pp. 543–559, DOI: 10.1007/978-3-642-33941-748.Google Scholar
2. Chen, G. Y., Gan, M., and Ding, F. (2019). “Modified Gram-Schmidt method based variable projection algorithm for separable nonlinear models.” IEEE Transactions on Neural Networks and Learning Systems, Vol. 30, No. 8, pp. 2410–2418, DOI: 10.1109/TNNLS.2018.2884909.
3. Chung, J. and Nagy, J. G (2010). “An efficient iterative approach for large-scale separable nonlinear inverse problems.” Siam Journal on Scientific Computing, Vol. 31, No. 6, pp. 4654–4674, DOI: 10.1137/080732213.
4. Gan, M. and Peng, H. (2012). “Stability analysis of RBF network-based statedependent autoregressive model for nonlinear time series.” Appl. Soft Comput, Vol. 12, No. 1, pp. 174–181, DOI: 10.1016/j.asoc.2011.08.055.
5. Golub, G. H. and Pereyra, V. (1973). “The differentiation of Pseudoinverses and nonlinear least squares problems whose variables separate.” Siam Journal on Numerical Analysis, Vol. 10, No. 2, pp. 413–432, DOI: 10.1137/0710036.
6. Guo, Y. Q., Tang, B., and Zhang, Q. J. (2012). “Research on coordinate transformation method of high accuracy based in space rectangular coordinates system.” Journal of Geodesy and Geodynamics, Vol. 32, No. 3, pp. 125–128.Google Scholar
7. Hong, J. H., Zach, C., and Fitzgibbon, A. (2017). Revisiting the variable projection method for separable nonlinear least squares problems, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5939–5947, Honolulu, HI, USA, DOI: 10.1109/CVPR.2017.629.Google Scholar
8. Kong, X. X. and Hu, M. L. (2017). “Improved algorithm of the morbidity problem of GM(1,1) model based on Cholesky and QR decomposition.” Chinese Journal of Management Science, Vol. 25, Special Issue, pp. 36–41.Google Scholar
9. Li, X. L. (2018). “Preconditioned stochastic gradient descent.” IEEE Transactions on Neural Networks and Learning Systems, Vol. 29, No. 5, pp. 1454–1466, DOI: 10.1109/TNNLS.2017.2672978.
10. Li, H. X. and Liu, Z. (2008). “A probabilistic neural-fuzzy learning system for stochastic modeling.” IEEE Trans. Fuzzy Syst., Vol. 16, No. 4, pp. 898–907, DOI: 10.1109/TFUZZ.2008.917302.
11. Moré, J. J. (1977). The Levenburg-Marquardt algorithm: Implementation and theory, Springer-Verlag, New York, NY, USA, pp. 105–116, DOI: 10.1007/BFb0067700.Google Scholar
12. Okatani, T. and Deguchi, K. (2007). “On the Wiberg algorithm for matrix factorization in the presence of missing components.” Int. J. Computer Vision, Vol. 72, pp. 329–337, DOI: 10.1007/s11263-006-9785-5.
13. Ruano, A. E., Jones, D. J., and Fleming, P. J. (1991). “A new formulation of the learning problem for a neural network controller.” Proc. 30th IEEE Conf. Decision Control, Brighton, UK, pp. 865–866, DOI:10.1109/CDC.1991.261439.Google Scholar
14. Shen, Y. Q. and Ypma, T. J. (1990). “Solving nonlinear systems of equations with only one nonlinear variable.” Journal of Computational & Applied Mathematics, Vol. 30, No. 2, pp. 235–246, DOI: 10.1016/0377-0427(90)90031-t.
15. Sjöberg, J. and Viberg, M. (1997). “Separable nonlinear least squares minimization—Possible improvements for neural net fitting.” Proc. IEEE Workshop Neural Netw., Signal Process, Amelia Island Plantation, FL, USA, pp. 345–354, DOI: 10.1109/NNSP.1997.622415.Google Scholar
16. Wang, H. C., Xu, J. J., and Wang, W. T. (2019). “Iteration algorithm of 3D Rectangular Coordinate Transformation of arbitrary rotation angle.” Journal of Geomatics, Vol. 44, No. 1, pp. 59–61, DOI: 10.14188/j.2095-6045.2017079.
17. Zhang, Y., Chai, T., and Wang, D. (2017). “An alternating identification algorithm for a class of nonlinear dynamical systems.” IEEE Transactions on Neural Networks and Learning Systems, Vol. 28, No. 7, pp. 1606–1617, DOI: 10.1109/TNNLS.2016.2547968.