{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,12]],"date-time":"2025-10-12T03:46:16Z","timestamp":1760240776537,"version":"build-2065373602"},"reference-count":50,"publisher":"MDPI AG","issue":"18","license":[{"start":{"date-parts":[[2019,9,16]],"date-time":"2019-09-16T00:00:00Z","timestamp":1568592000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>The concept presented in this paper is based on previous dynamical methods to realize a time-varying matrix inversion. It is essentially a set of coupled ordinary differential equations (ODEs) which does indeed constitute a recurrent neural network (RNN) model. The coupled ODEs constitute a universal modeling framework for realizing a matrix inversion provided the matrix is invertible. The proposed model does converge to the inverted matrix if the matrix is invertible, otherwise it converges to an approximated inverse. Although various methods exist to solve a matrix inversion in various areas of science and engineering, most of them do assume that either the time-varying matrix inversion is free of noise or they involve a denoising module before starting the matrix inversion computation. However, in the practice, the noise presence issue is a very serious problem. Also, the denoising process is computationally expensive and can lead to a violation of the real-time property of the system. Hence, the search for a new \u2018matrix inversion\u2019 solving method inherently integrating noise-cancelling is highly demanded. In this paper, a new combined\/extended method for time-varying matrix inversion is proposed and investigated. The proposed method is extending both the gradient neural network (GNN) and the Zhang neural network (ZNN) concepts. Our new model has proven that it has exponential stability according to Lyapunov theory. Furthermore, when compared to the other previous related methods (namely GNN, ZNN, Chen neural network, and integration-enhanced Zhang neural network or IEZNN) it has a much better theoretical convergence speed. To finish, all named models (the new one versus the old ones) are compared through practical examples and both their respective convergence and error rates are measured. It is shown\/observed that the novel\/proposed method has a better practical convergence rate when compared to the other models. Regarding the amount of noise, it is proven that there is a very good approximation of the matrix inverse even in the presence of noise.<\/jats:p>","DOI":"10.3390\/s19184002","type":"journal-article","created":{"date-parts":[[2019,9,17]],"date-time":"2019-09-17T03:31:46Z","timestamp":1568691106000},"page":"4002","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":10,"title":["A Novel Recurrent Neural Network-Based Ultra-Fast, Robust, and Scalable Solver for Inverting a \u201cTime-Varying Matrix\u201d"],"prefix":"10.3390","volume":"19","author":[{"given":"Vahid","family":"Tavakkoli","sequence":"first","affiliation":[{"name":"Institute for Smart Systems Technologies, University Klagenfurt, A9020 Klagenfurt, Austria"}]},{"given":"Jean Chamberlain","family":"Chedjou","sequence":"additional","affiliation":[{"name":"Institute for Smart Systems Technologies, University Klagenfurt, A9020 Klagenfurt, Austria"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0773-9476","authenticated-orcid":false,"given":"Kyandoghere","family":"Kyamakya","sequence":"additional","affiliation":[{"name":"Institute for Smart Systems Technologies, University Klagenfurt, A9020 Klagenfurt, Austria"}]}],"member":"1968","published-online":{"date-parts":[[2019,9,16]]},"reference":[{"key":"ref_1","unstructured":"Lumpkin, B. (1997). Algebra Activities from Many Cultures, J. Weston Walch."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"414","DOI":"10.1109\/TEVC.2014.2336865","article-title":"Locating Multiple Optimal Solutions of Nonlinear Equation Systems Based on Multiobjective Optimization","volume":"19","author":"Song","year":"2015","journal-title":"IEEE Trans. Evol. Comput."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"389","DOI":"10.1109\/LCOMM.2013.010313.121837","article-title":"Sphere Decoding for MIMO Systems with Newton Iterative Matrix Inversion","volume":"17","author":"Wang","year":"2013","journal-title":"IEEE Commun."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"1304","DOI":"10.1109\/TNNLS.2013.2250300","article-title":"Feasibility and Finite Convergence Analysis for Accurate On-Line \u03bd-Support Vector Machine","volume":"24","author":"Gu","year":"2013","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Eilert, J., Wu, D., and Liu, D. (2007, January 27\u201330). Efficient Complex Matrix Inversion for MIMO Software Defined Radio. Proceedings of the IEEE International Symposium on Circuits and Systems, New Orleans, LA, USA.","DOI":"10.1109\/ISCAS.2007.377850"},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Wu, M., Yin, B., Vosoughi, A., Studer, C., Cavallaro, J.R., and Dick, C. (2013, January 19\u201323). Approximate matrix inversion for high-throughput data detection in the large-scale MIMO uplink. Proceedings of the ISCAS2013, Beijing, China.","DOI":"10.1109\/ISCAS.2013.6572301"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1858","DOI":"10.1109\/TSP.2011.2105485","article-title":"QR Decomposition-Based Matrix Inversion for High Performance Embedded MIMO Receivers","volume":"59","author":"Ma","year":"2011","journal-title":"IEEE Trans. Signal Process."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Arias-Garc\u00eda, J., Jacobi, R.P., Llanos, C.H., and Ayala-Rinc\u00f3n, M. (2011, January 13\u201315). A suitable FPGA implementation of floating-point matrix inversion based on Gauss-Jordan elimination. Proceedings of the VII Southern Conference on Programmable Logic (SPL), Cordoba, Argentina.","DOI":"10.1109\/SPL.2011.5782659"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Irturk, A., Benson, B., Mirzaei, S., and Kastner, R. (2008, January 8\u20139). An FPGA Design Space Exploration Tool for Matrix Inversion Architectures. Proceedings of the Symposium on Application Specific Processors, Anaheim, CA, USA.","DOI":"10.1109\/SASP.2008.4570784"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"384","DOI":"10.1121\/1.428310","article-title":"Adaptive eigenvalue decomposition algorithm for passive acoustic source localization","volume":"107","author":"Benesty","year":"2001","journal-title":"J. Acoust. Soc. Am."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Warp, R.J., Godfrey, D., and Dobbins, J.T. (2000). Applications of matrix inversion tomosynthesis. Med. Imaging, 3977.","DOI":"10.1117\/12.384512"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"021907","DOI":"10.1118\/1.4773891","article-title":"The effect of averaging adjacent planes for artifact reduction in matrix inversion tomosynthesis","volume":"40","author":"Godfrey","year":"2013","journal-title":"Med Phys."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1477","DOI":"10.1109\/TNN.2005.857946","article-title":"Design and analysis of a general recurrent neural network model for time-varying matrix inversion","volume":"16","author":"Zhang","year":"2005","journal-title":"IEEE Trans. Neural Netw."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"22","DOI":"10.1016\/j.neucom.2012.05.012","article-title":"Zhang neural network, Getz\u2013Marsden dynamic system, and discrete-time algorithms for time-varying matrix inversion with application to robots\u2019 kinematic control","volume":"97","author":"Guo","year":"2012","journal-title":"Neurocomputing"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Guo, D., Li, K., Yan, L., Nie, Z., and Jin, F. (2016, January 28\u201330). The application of Li-function activated RNN to acceleration-level robots\u2019 kinematic control via time-varying matrix inversion. Proceedings of the Chinese Control and Decision Conference (CCDC), Yinchuan, China.","DOI":"10.1109\/CCDC.2016.7531580"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Amato, F., Moscato, V., Picariello, A., and Sperl\u00ed, G. (2017, January 19\u201321). Recommendation in Social Media Networks. Proceedings of the IEEE Third International Conference on Multimedia Big Data (BigMM), Laguna Hills, CA, USA.","DOI":"10.1109\/BigMM.2017.55"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Amato, F., Castiglione, A., Moscato, V., Picariello, A., and Sperli, G. (2018). Multimedia summarization using social media content. Multimed. Tools Appl., 77.","DOI":"10.1007\/s11042-017-5556-2"},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"12092","DOI":"10.1364\/OE.26.012092","article-title":"Iterative matrix algorithm for high precision temperature and force decoupling in multi-parameter FBG sensing","volume":"26","author":"Hopf","year":"2018","journal-title":"Opt. Express"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"864","DOI":"10.1109\/TBCAS.2016.2539244","article-title":"Matrix-Inversion-Free Compressed Sensing with Variable Orthogonal Multi-Matching Pursuit Based on Prior Information for ECG Signals","volume":"10","author":"Cheng","year":"2016","journal-title":"IEEE Trans. Biomed. Circuits Syst."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1109\/TSP.2008.2005866","article-title":"Multitask Compressive Sensing","volume":"57","author":"Ji","year":"2009","journal-title":"IEEE Trans. Signal Process."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"319","DOI":"10.1088\/0957-0233\/5\/4\/001","article-title":"Optimal design of multivariate sensors","volume":"5","author":"Bicchi","year":"1994","journal-title":"Meas. Sci. Technol."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"1576","DOI":"10.1175\/JTECH2080.1","article-title":"General matrix inversion technique for the calibration of electric field sensor arrays on aircraft platforms","volume":"24","author":"Mach","year":"2007","journal-title":"J. Atmos. Ocean. Technol."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"3509","DOI":"10.1109\/TSP.2016.2550005","article-title":"Sensor Selection for Estimation with Correlated Measurement Noise","volume":"64","author":"Liu","year":"2016","journal-title":"IEEE Trans. Signal Process."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"1940","DOI":"10.1109\/TAC.2009.2023779","article-title":"Performance Analysis of Gradient Neural Network Exploited for Online Time-Varying Matrix Inversion","volume":"54","author":"Zhang","year":"2009","journal-title":"IEEE Trans. Autom. Control"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"1405","DOI":"10.1109\/TCSI.2008.2007065","article-title":"From Zhang Neural Network to Newton Iteration for Matrix Inversion","volume":"56","author":"Zhang","year":"2009","journal-title":"IEEE Trans. Circuits Syst."},{"key":"ref_26","unstructured":"Kandasamy, W., and Smarandache, F. (2012). Exploring the Extension of Natural Operations on Intervals, Matrices and Complex Numbers, ZIP Publishing."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"876","DOI":"10.1016\/j.ipl.2013.09.002","article-title":"Improved neural solution for the Lyapunov matrix equation based on gradient search","volume":"113","author":"Chen","year":"2013","journal-title":"Inf. Process. Lett."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"780","DOI":"10.1016\/j.ipl.2011.05.010","article-title":"Improved gradient-based neural networks for online solution of Lyapunov matrix equation","volume":"111","author":"Yi","year":"2011","journal-title":"Inf. Process. Lett."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"281","DOI":"10.1145\/321075.321076","article-title":"Error Analysis of Direct Methods of Matrix Inversion","volume":"8","author":"Wilkinson","year":"1961","journal-title":"JACM"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"1257","DOI":"10.1109\/31.7600","article-title":"Cellular Neural Networks: Theory","volume":"35","author":"Chua","year":"1988","journal-title":"IEEE Trans. Circuits Syst."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"163","DOI":"10.1109\/82.222815","article-title":"The CNN universal machine: An analogic array computer","volume":"40","author":"Roska","year":"1993","journal-title":"IEEE Trans. Circuits Syst."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"540","DOI":"10.1109\/TNN.2009.2013243","article-title":"Comments on Backpropagation Algorithms for a Broad Class of Dynamic Networks","volume":"20","author":"Endisch","year":"2009","journal-title":"IEEE Trans. Neural Netw."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Potluri, S., Fasih, A., Kishore, L., Machot, F.A., and Kyamakya, K. (2011, January 25\u201327). CNN Based High Performance Computing for Real Time Image Processing on GPU. Proceedings of the Nonlinear Dynamics and Synchronization (INDS) & 16th Int\u2019l Symposium on Theoretical Electrical Engineering (ISTET), Klagenfurt, Austria.","DOI":"10.1109\/INDS.2011.6024781"},{"key":"ref_34","unstructured":"Mainzer, K. (2002, January 24). CNN and the Evoulution of Complex Information Systems in Nature and Technology. Proceedings of the 7th IEEE Cellular Neural Networks and Their Applications, Frankfurt, Germany."},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Chedjou, J., Kyamakya, K., Khan, U., and Latif, M. (2010, January 3\u20135). Potential Contribution of CNN-based Solving of Stiff ODEs & PDEs to Enabling Real-Time Computational Engineering. Proceedings of the 2010 12th International Workshop on Cellular Nanoscale Networks and their Applications (CNNA 2010), Berkeley, CA, USA.","DOI":"10.1109\/CNNA.2010.5430262"},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"749","DOI":"10.1109\/TNNLS.2014.2323218","article-title":"A Universal Concept Based on Cellular Neural Networks for Ultrafast and Flexible Solving of Differential Equations","volume":"26","author":"Chedjou","year":"2015","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"2830","DOI":"10.1109\/TNNLS.2015.2397551","article-title":"Recurrent Neural Network for Computing the Drazin Inverse","volume":"26","author":"Stanimirovic","year":"2015","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"1479","DOI":"10.1137\/S1064827594267161","article-title":"Recurrent Neural Networks for Computing Pseudoinverses of Rank-Deficient Matrices","volume":"5","author":"Wang","year":"1997","journal-title":"Siam J. Sci. Comput."},{"key":"ref_39","first-page":"10218","article-title":"Recurrent Implicit Dynamics for Online Matrix Inversion","volume":"219","author":"Chen","year":"2013","journal-title":"Appl. Math. Comput."},{"key":"ref_40","unstructured":"Feng, F., Zhang, Q., and Liu, H. (2015, January 20\u201323). A Recurrent Neural Network for Extreme Eigenvalue Problem. Proceedings of the ICIC 2005, Fuzhou, China."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"1385","DOI":"10.1016\/j.camwa.2010.06.020","article-title":"Another neural network based approach for commuting eigenvalues and eigenvectors of real skew-symmetric matrices","volume":"60","author":"Tang","year":"2010","journal-title":"Comput. Math. Appl."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"812","DOI":"10.1109\/3477.956043","article-title":"A PCA approach for fast retrieval of structural patterns in attributed graphs","volume":"31","author":"Xu","year":"2001","journal-title":"IEEE Trans. Syst. Man Cybern."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1016\/j.camwa.2006.09.004","article-title":"A recurrent neural network computing the largest imaginary or real part of eigenvalues of real matrices","volume":"53","author":"Liu","year":"2007","journal-title":"Comput. Math. Appl."},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"293","DOI":"10.1109\/72.207617","article-title":"Neural Network for Quadratic Optimization with Bound Constraints","volume":"4","author":"Bouzerdorm","year":"1993","journal-title":"IEEE Trans. Neural Netw."},{"key":"ref_45","unstructured":"Zhang, Y. (2006, January 23\u201325). Towards piecewise-linear primal neural networks for optimization and redundant robotics. Proceedings of the IEEE International Conference on Networking, Sensing and Control, Ft. Lauderdale, FL, USA."},{"key":"ref_46","first-page":"141","article-title":"Neural Computation of Decisions in Optimization Problems","volume":"52","author":"Hopfield","year":"1984","journal-title":"Cybernetics"},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"554","DOI":"10.1109\/31.1783","article-title":"Neural Networks for Nonlinear Programming","volume":"35","author":"Kenndey","year":"1988","journal-title":"IEEE Trans Circuits Syst."},{"key":"ref_48","unstructured":"He, Z., Gao, S., Xiao, L., Liu, D., He, H., and Barber, D. (2017, January 4\u20139). Wider and deeper, cheaper and faster: Tensorized LSTMs for sequence learning. Proceedings of the Conference on Neural Information Processing Systems 2017, Long Beach, CA, USA."},{"key":"ref_49","unstructured":"Jang, J.-S., Lee, S.-Y., and Shin, S.-Y. (1987, January 8\u201312). An Optimization Network for Matrix Inversion. Proceedings of the 1987 IEEE Conference on Neural Information Processing Systems, Denver, CO, USA."},{"key":"ref_50","first-page":"10066","article-title":"Complex-valued Zhang neural network for online complex-valued time-varying matrix inversion","volume":"217","author":"Zhang","year":"2011","journal-title":"Appl. Math. Comput."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/19\/18\/4002\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T13:20:41Z","timestamp":1760188841000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/19\/18\/4002"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2019,9,16]]},"references-count":50,"journal-issue":{"issue":"18","published-online":{"date-parts":[[2019,9]]}},"alternative-id":["s19184002"],"URL":"https:\/\/doi.org\/10.3390\/s19184002","relation":{},"ISSN":["1424-8220"],"issn-type":[{"type":"electronic","value":"1424-8220"}],"subject":[],"published":{"date-parts":[[2019,9,16]]}}}