Robust machine learning modeling for predictive control using Lipschitz-Constrained Neural Networks

被引:4
作者
Tan, Wallace Gian Yion [1 ]
Wu, Zhe [1 ]
机构
[1] Natl Univ Singapore, Dept Chem & Biomol Engn, Singapore 117585, Singapore
关键词
Lipschitz-Constrained Neural Networks; Robust machine learning model; Generalization error; Model predictive control; Neural network sensitivity; Over-fitting; DROPOUT;
D O I
10.1016/j.compchemeng.2023.108466
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Neural networks (NNs) have emerged as a state-of-the-art method for modeling nonlinear systems in model predictive control (MPC). However, the robustness of NNs, in terms of sensitivity to small input perturbations, remains a critical challenge for practical applications. To address this, we develop Lipschitz-Constrained Neural Networks (LCNNs) for modeling nonlinear systems and derive rigorous theoretical results to demonstrate their effectiveness in approximating Lipschitz functions, reducing input sensitivity, and preventing over-fitting. Specifically, we first prove a universal approximation theorem to show that LCNNs using SpectralDense layers can approximate any 1-Lipschitz target function. Then, we prove a probabilistic generalization error bound for LCNNs using SpectralDense layers by using their empirical Rademacher complexity. Finally, the LCNNs are incorporated into the MPC scheme, and a chemical process example is utilized to show that LCNN-based MPC outperforms MPC using conventional feedforward NNs in the presence of training data noise.
引用
收藏
页数:14
相关论文
共 32 条
[1]  
Anil C, 2019, PR MACH LEARN RES, V97
[2]  
[Anonymous], 2018, PROC INT C LEARN TH
[3]   Perturbation Analysis of Learning Algorithms: Generation of Adversarial Examples From Classification to Regression [J].
Balda, Emilio Rafael ;
Behboodi, Arash ;
Mathar, Rudolf .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2019, 67 (23) :6078-6091
[4]  
Baldi P., 2013, Advances in Neural Information Processing Systems, P2814
[5]   The dropout learning algorithm [J].
Baldi, Pierre ;
Sadowski, Peter .
ARTIFICIAL INTELLIGENCE, 2014, 210 :78-122
[6]   LipBaB: Computing Exact Lipschitz Constant of ReLU Networks [J].
Bhowmick, Aritra ;
D'Souza, Meenakshi ;
Raghavan, G. Srinivasa .
ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING - ICANN 2021, PT IV, 2021, 12894 :151-162
[7]  
Chen H., 1995, Proceedings of the Third European Control Conference. ECC 95, P3247
[8]  
Cortes C, 2012, Arxiv, DOI [arXiv:1205.2653, DOI 10.48550/ARXIV.1205.2653]
[9]  
DeNero J, 2011, S MACH LEARN SPEECH
[10]  
Federer H., 1969, GEOMETRIC MEASURE TH