Large Language Model-Driven Evolutionary Optimization With a Hallucination-Resilient Mechanism

Weixiong Huang; Rui Wang; Tao Zhang; Sheng Qi; Shichao Fan; Ling Wang

doi:10.1109/JAS.2025.125876

Volume 13 Issue 6

Jun. 2026

IEEE/CAA Journal of Automatica Sinica

JCR Impact Factor: 18.3, Top 1 (SCI Q1)

CiteScore: 28.2, Top 1% (Q1)
Google Scholar h5-index: 95， TOP 5

Turn off MathJax

Article Contents

Article Navigation > IEEE/CAA Journal of Automatica Sinica > 2026 > 13(6): 1428-1445

W. Huang, R. Wang, T. Zhang, S. Qi, S. Fan, and L. Wang, “Large language model-driven evolutionary optimization with a hallucination-resilient mechanism,” IEEE/CAA J. Autom. Sinica, vol. 13, no. 6, pp. 1428–1445, Jun. 2026. doi: 10.1109/JAS.2025.125876

Citation:

W. Huang, R. Wang, T. Zhang, S. Qi, S. Fan, and L. Wang, “Large language model-driven evolutionary optimization with a hallucination-resilient mechanism,” IEEE/CAA J. Autom. Sinica, vol. 13, no. 6, pp. 1428–1445, Jun. 2026. doi: 10.1109/JAS.2025.125876

Citation:

PDF( 3835 KB)

Large Language Model-Driven Evolutionary Optimization With a Hallucination-Resilient Mechanism

doi: 10.1109/JAS.2025.125876

Funds: This work was supported by the National Natural Science Foundation of China (62550020, 72421002), the Science and Technology Project for Young and Middle-aged Talents of Hunan (2023TJ-Z03), the University Fundamental Research Fund (23-ZZCX-JDZ-28), and the National Postdoctoral Program for Innovative Talents of China (BX20250439). The authors would also like to thank the support from COSTA: Complex System Optimization Team of the College of System Engineering at NUDT

More Information

Author Bio:
Weixiong Huang received the M.Sc. degree from the College of Computer Science and Cyberspace Security, Xiangtan University in 2022. He is currently pursuing the Ph.D. degree from the College of Systems Engineering, National University of Defense Technology. His current research interests include multiobjective optimization, evolutionary computation, and large language models for optimization

Rui Wang (Senior Member, IEEE) received the B.S. degree from National University of Defense Technology (NUDT) in 2008, and the Ph.D. degree from the University of Sheffield, U.K., in 2013. He is currently with the National University of Defense Technology. His current research interests include multiobjective optimization and the development of algorithms applicable in practice. He has authored more than 40 referred papers including those published in IEEE Transactions on Evolutionary Computation and IEEE Transactions on Cybernetics. Dr. Wang serves as an Associate Editors of the IEEE Transactions on Evolutionary Computation, Swarm and Evolutionary Computation, Expert System With Applications, etc. He is the recipients of The Operational Research Society Ph.D. Prize in 2014, of the Funds for Distinguished Young Scientists from the Natural Science Foundation of Hunan province at 2016, of the Wu Wen-Jun Artificial Intelligence Outstanding Young Scholar at 2017, of the National Science Fund for Outstanding Young Scholars at 2021

Tao Zhang received the B.S., M.S., Ph.D. degrees from National University of Defense Technology (NUDT), in 1998, 2001, and 2004, respectively.He is a Full Professor with the College of System Engineering, NUDT. He is also the Director of the Hunan Key Laboratory of Multi-Energy System Intelligent Interconnection Technology (HKL-MSI2T). His current research interests include optimal scheduling, data mining, and optimization methods on energy internet. He was the recipient of the Science and Technology Award of Provincial Level (First Place in 2020, 2021, and Second Place in 2015, 2018)

Sheng Qi received the M.Sc. degree from the College of Computer Science and Cyberspace Security, Xiangtan University in 2022. He is currently pursuing the Ph.D. degree from the College of Systems Engineering, National University of Defense Technology. His current research interests include evolutionary computation, multiobjective optimization, and optimization methods on energy internet network

Shichao Fan received the B.Sc. degree from the National University of Defense Technology in 2008, and the M.Sc. degree from Central South University in 2014. His current research interests include evolutionary computation, multiobjective optimization, and engineering management scheduling optimization

Ling Wang received the B.Sc. in automation and the Ph.D. degree in control theory and control engineering from Tsinghua University in 1995 and 1999, respectively. Since 1999, he has been with the Department of Automation, Tsinghua University, where he became a Full Professor in 2008. His current research interests include intelligent optimization and production scheduling. He was the recipient of the National Natural Science Fund for Distinguished Young Scholars of China, the National Natural Science Award (Second Place) in 2014, the Science and Technology Award of Beijing City in 2008, and the Natural Science Award (First Place in 2003, and Second Place in 2007) nominated by the Ministry of Education of China
Corresponding author: Rui Wang, e-mail: rui_wang@nudt.edu.cn
Received Date: 2025-03-01
Revised Date: 2025-06-04
Accepted Date: 2025-08-19

Abstract

Abstract

Large language models (LLMs) have demonstrated significant potential as black-box optimizers due to their strong reasoning capabilities. However, challenges such as the hallucination phenomenon introduce instability and uncertainty, limiting their effectiveness. This paper proposes an LLM-driven evolutionary optimization framework, referred to as LLM-driven hybrid evolutionary optimization framework (LHO), that integrates LLMs with traditional evolutionary operators. LLMs accelerate the optimization process by generating high-quality solutions, while evolutionary operators ensure stability and provide performance guarantees. To further enhance robustness, we introduce a hallucination-resilient mechanism to mitigate the risks associated with LLM hallucinations. Experimental results on various benchmark tests, encompassing single-objective, multiobjective, and complex constrained multiobjective problems, confirm the effectiveness and practicality of the proposed framework, offering valuable insights and future directions for LLM as evolutionary optimizers.
- Evolutionary algorithm (EA),
- evolutionary computation,
- large language models (LLMs)

FullText(HTML)

References(59)

References

[1]	Z. Wang, H.-L. Zhen, J. Deng, Q. Zhang, X. Li, M. Yuan, and J. Zeng, “Multiobjective optimization-aided decision-making system for large-scale manufacturing planning,” IEEE Trans. Cybern., vol. 52, no. 8, pp. 8326–8339, Aug. 2022. doi: 10.1109/TCYB.2021.3049712
[2]	R. Espinosa, F. Jiménez, and J. Palma, “Multi-surrogate assisted multi-objective evolutionary algorithms for feature selection in regression and classification problems with time series data,” Inf. Sci., vol. 622, pp. 1064–1091, Apr. 2023. doi: 10.1016/j.ins.2022.12.004
[3]	Y. Chen, R. Wang, M. Ming, S. Cheng, Y. Bao, W. Zhang, and C. Zhang, “Constraint multi-objective optimal design of hybrid renewable energy system considering load characteristics,” Complex Intell. Syst., vol. 8, no. 2, pp. 803–817, Apr. 2022. doi: 10.1007/s40747-021-00363-4
[4]	C. A. Coello Coello, “Constraint-handling techniques used with evolutionary algorithms,” in Proc. Genetic and Evolutionary Computation Conf. Companion, Melbourne, Australia, 2024, pp. 1261−1283.
[5]	J. H. Holland, Adaptation in Natural and Artificial Systems: An Introductory Analysis With Applications to Biology, Control, and Artificial Intelligence. Cambridge, UAS: MIT Press, 1992.
[6]	R. Storn and K. Price, “Differential evolution-A simple and efficient heuristic for global optimization over continuous spaces,” J. Glob. Optim., vol. 11, no. 4, pp. 341–359, Dec. 1997. doi: 10.1023/A:1008202821328
[7]	R. Eberhart and J. Kennedy, “A new optimizer using particle swarm theory,” in Proc. 6th Int. Symp. Micro Machine and Human Science, Nagoya, Japan, 1995, pp. 39−43.
[8]	K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, “A fast and elitist multiobjective genetic algorithm: NSGA-II,” IEEE Trans. Evol. Comput., vol. 6, no. 2, pp. 182–197, Apr. 2002. doi: 10.1109/4235.996017
[9]	E. Zitzler, M. Laumanns, and L. Thiele, “SPEA2: Improving the strength pareto evolutionary algorithm,” ETH Zurich, Computer Engineering and Networks Laboratory, Zurich, Switzerland, TIK-Rep. 103, 2001.
[10]	Q. Zhang and H. Li, “MOEA/D: A multiobjective evolutionary algorithm based on decomposition,” IEEE Trans. Evol. Comput., vol. 11, no. 6, pp. 712–731, Dec. 2007. doi: 10.1109/TEVC.2007.892759
[11]	J. Bader and E. Zitzler, “HypE: An algorithm for fast hypervolume-based many-objective optimization,” Evol. Comput., vol. 19, no. 1, pp. 45–76, Mar. 2011. doi: 10.1162/EVCO_a_00009
[12]	K. Sanderson, “GPT-4 is here: What scientists think,” Nature, vol. 615, no. 7954, Art. no. 773, Mar. 2023. doi: 10.1038/d41586-023-00816-5
[13]	C. Yang, X. Wang, Y. Lu, H. Liu, Q. V. Le, D. Zhou, and X. Chen, “Large language models as optimizers,” Proc. 12th Int. Conf. Learning Representations, Vienna, Austria, pp. 1–41, 2024.
[14]	P.-F. Guo, Y.-H. Chen, Y.-D. Tsai, and S.-D. Lin, “Towards optimizing with large language models,” arXiv preprint arXiv: 2310.05204, 2023.
[15]	S. Yao, D. Yu, J. Zhao, I. Shafran, T. L. Griffiths, Y. Cao, and K. Narasimhan, “Tree of thoughts: Deliberate problem solving with large language models,” Proc. 37th Int. Conf. Neural Information Processing Systems, New Orleans, USA, Art. no. 517, 2023.
[16]	T. Kojima, S. S. Gu, M. Reid, Y. Matsuo, and Y. Iwasawa, “Large language models are zero-shot reasoners,” in Proc. 36th Int. Conf. Neural Information Processing Systems, New Orleans, USA, 2022, Art. no. 1613.
[17]	B. Min, H. Ross, E. Sulem, A. P. B. Veyseh, T. H. Nguyen, O. Sainz, E. Agirre, I. Heintz, and D. Roth, “Recent advances in natural language processing via large pre-trained language models: A survey,” ACM Comput. Surv., vol. 56, no. 2, Art. no. 30, Feb. 2024.
[18]	A. Renda, A. Hopkins, and M. Carbin, “Can LLMs generate random numbers? Evaluating LLM sampling in controlled domains,” in Proc. Sampling and Optimization in Discrete Space, Honolulu, USA, 2023, pp. 1−22.
[19]	Y. Zhou and J. He, “A runtime analysis of evolutionary algorithms for constrained optimization problems,” IEEE Trans. Evol. Comput., vol. 11, no. 5, pp. 608–619, Oct. 2007. doi: 10.1109/TEVC.2006.888929
[20]	Y. Diouane, S. Gratton, and L. N. Vicente, “Globally convergent evolution strategies,” Math. Program., vol. 152, no. 1−2, pp. 467–490, Aug. 2015. doi: 10.1007/s10107-014-0793-x
[21]	C. A. Coello Coello, “Theoretical and numerical constraint-handling techniques used with evolutionary algorithms: A survey of the state of the art,” Comput. Methods Appl. Mech. Eng., vol. 191, no. 11−12, pp. 1245–1287, Jan. 2002. doi: 10.1016/S0045-7825(01)00323-1
[22]	K. M. Sallam, S. M. Elsayed, R. K. Chakrabortty, and M. J. Ryan, “Improved multi-operator differential evolution algorithm for solving unconstrained problems,” in Proc. IEEE Congr. Evolutionary Computation, Glasgow, UK, 2020, pp. 1−8.
[23]	R. Cheng and Y. Jin, “A competitive swarm optimizer for large scale optimization,” IEEE Trans. Cybern., vol. 45, no. 2, pp. 191–204, Feb. 2015. doi: 10.1109/TCYB.2014.2322602
[24]	K. M. Sallam, S. M. Elsayed, R. A. Sarker, and D. L. Essam, “Landscape-based adaptive operator selection mechanism for differential evolution,” Inf. Sci., vol. 418−419, pp. 383–404, Dec. 2017. doi: 10.1016/j.ins.2017.08.028
[25]	C. A. Coello Coello, “Evolutionary multi-objective optimization: A historical view of the field,” IEEE Comput. Intell. Mag., vol. 1, no. 1, pp. 28–36, Feb. 2006. doi: 10.1109/MCI.2006.1597059
[26]	Y. Xiang, Y. Zhou, M. Li, and Z. Chen, “A vector angle-based evolutionary algorithm for unconstrained many-objective optimization,” IEEE Trans. Evol. Comput., vol. 21, no. 1, pp. 131–152, Feb. 2017. doi: 10.1109/TEVC.2016.2587808
[27]	K. Deb and H. Jain, “An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, Part I: Solving problems with box constraints,” IEEE Trans. Evol. Comput., vol. 18, no. 4, pp. 577–601, Aug. 2014. doi: 10.1109/TEVC.2013.2281535
[28]	E. Zitzler and S. Künzli, “Indicator-based selection in multiobjective search,” in Proc. 8th Int. Conf. Parallel Problem Solving From Nature, Birmingham, UK, 2004, pp. 832−842.
[29]	F. Liu, X. Lin, S. Yao, Z. Wang, X. Tong, M. Yuan, and Q. Zhang, “Large language model for multiobjective evolutionary optimization,” in Proc. 13th Int. Conf. Evolutionary Multi-Criterion Optimization, Canberra, Australia, 2025, pp. 178−191.
[30]	S. Liu, C. Chen, X. Qu, K. Tang, and Y.-S. Ong, “Large language models as evolutionary optimizers,” in Proc. IEEE Congr. Evolutionary Computation, Yokohama, Japan, 2024, pp. 1−8.
[31]	S. Brahmachary, S. M. Joshi, A. Panda, K. Koneripalli, A. K. Sagotra, H. Patel, A. Sharma, A. D. Jagtap, and K. Kalyanaraman, “Large language model-based evolutionary optimizer: Reasoning with elitism,” Neurocomputing, vol. 622, Art. no. 129272, Mar. 2025. doi: 10.1016/j.neucom.2024.129272
[32]	H. Bradley, A. Dai, H. Teufel, J. Zhang, K. Oostermeijer, M. Bellagente, J. Clune, K. Stanley, G. Schott, and J. Lehman, “Quality-diversity through AI feedback,” in Proc. 12th Int. Conf. Learning Representations, Vienna, Austria, 2024, 1−112.
[33]	R. Lange, Y. Tian, and Y. Tang, “Large language models as evolution strategies,” in Proc. Genetic and Evolutionary Computation Conf. Companion, Melbourne, Australia, 2024, pp. 579−582.
[34]	L. L. Custode, F. Caraffini, A. Yaman, and G. Iacca, “An investigation on the use of large language models for hyperparameter tuning in evolutionary algorithms,” in Proc. Genetic and Evolutionary Computation Conf. Companion, Melbourne, Australia, 2024, pp. 1838−1845.
[35]	F. Liu, X. Tong, M. Yuan, X. Lin, F. Luo, Z. Wang, Z. Lu, and Q. Zhang, “Evolution of heuristics: Towards efficient automatic algorithm design using large language model,” in Proc. 41st Int. Conf. Machine Learning, Vienna, Austria, 2024, pp. 32201−32223.
[36]	B. Romera-Paredes, M. Barekatain, A. Novikov, M. Balog, M. P. Kumar, E. Dupont, et al, “Mathematical discoveries from program search with large language models,” Nature, vol. 625, no. 7995, pp. 468–475, Jan. 2024. doi: 10.1038/s41586-023-06924-6
[37]	F. Liu, X. Tong, M. Yuan, and Q. Zhang, “Algorithm evolution using large language model,” arXiv preprint arXiv: 2311.15249, 2023.
[38]	S. Yao, F. Liu, X. Lin, Z. Lu, Z. Wang, and Q. Zhang, “Multi-objective evolution of heuristic using large language model,” in Proc. 39th AAAI Conf. Artificial Intelligence, Philadelphia, USA, 2025, 27144−27152.
[39]	H. Ye, J. Wang, Z. Cao, F. Berto, C. Hua, H. Kim, J. Park, and G. Song, “ReEvo: Large language models as hyper-heuristics with reflective evolution,” in Proc. 38th Int. Conf. Neural Information Processing Systems, Vancouver, Canada, 2024, Art. no. 1381.
[40]	Y. J. Ma, W. Liang, G. Wang, D.-A. Huang, O. Bastani, D. Jayaraman, Y. Zhu, L. Fan, and A. Anandkumar, “Eureka: Human-level reward design via coding large language models,” in Proc. 12th Int. Conf. Learning Representations, Vienna, Austria, 2024, 1−45.
[41]	N. van Stein and T. Bäck, “LLaMEA: A large language model evolutionary algorithm for automatically generating metaheuristics,” IEEE Trans. Evol. Comput., vol. 29, no. 2, pp. 331–345, Apr. 2025. doi: 10.1109/TEVC.2024.3497793
[42]	OpenAI, “GPT-3.5 turbo,” version: 0125 [Online]. Available: https://developers.openai.com/api/docs/models/gpt-3.5-turbo, Accessed on: Nov. 20, 2024.
[43]	OpenAI, “GPT-4 turbo,” version: 2024-04-09 [Online]. Available: https://developers.openai.com/api/docs/models/gpt-4-turbo, Accessed on: Dec. 1, 2024.
[44]	OpenAI, “GPT-4o,” version: 2024-08-06 [Online]. Available: https://developers.openai.com/api/docs/models/gpt-4o, Accessed on: Dec. 3, 2024.
[45]	OpenAI, “GPT-4o mini,” version: 2024-07-18 [Online]. Available: https://developers.openai.com/api/docs/models/gpt-4o-mini, Accessed on: Dec. 3, 2024.
[46]	OpenAI, “ChatGPT-o1,” version: 2024-12-17 [Online]. Available: https://developers.openai.com/api/docs/models/gpt-o1, Accessed on: Dec. 20, 2024.
[47]	OpenAI, “ChatGPT-o1-mini,” version: 2024-09-12 [Online]. Available: https://developers.openai.com/api/docs/models/gpt-o1-mini, Accessed on: Dec. 20, 2024.
[48]	T. B. Brown, B. Mann, N. Ryder, M. Subbiah, J. Kaplan, P. Dhariwal, et al., “Language models are few-shot learners,” in Proc. 34th Int. Conf. Neural Information Processing Systems, Vancouver, Canada, 2020, Art. no. 159.
[49]	J. Wei, X. Wang, D. Schuurmans, M. Bosma, B. Ichter, F. Xia, E. H. Chi, Q. V. Le, and D. Zhou, “Chain-of-thought prompting elicits reasoning in large language models,” in Proc. 36th Int. Conf. Neural Information Processing Systems, New Orleans, USA, 2022, Art. no. 1800.
[50]	X. Wang, J. Wei, D. Schuurmans, Q. V. Le, E. H. Chi, S. Narang, A. Chowdhery, and D. Zhou, “Self-consistency improves chain of thought reasoning in language models,” in Proc. 11th Int. Conf. Learning Representations, Kigali, Rwanda, 2023, pp. 12−24.
[51]	X. Yao, Y. Liu, and G. Lin, “Evolutionary programming made faster,” IEEE Trans. Evol. Comput., vol. 3, no. 2, pp. 82–102, Jul. 1999. doi: 10.1109/4235.771163
[52]	E. Zitzler, K. Deb, and L. Thiele, “Comparison of multiobjective evolutionary algorithms: Empirical results,” Evol. Comput., vol. 8, no. 2, pp. 173–195, Jun. 2000. doi: 10.1162/106365600568202
[53]	Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, “An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions,” Soft Comput., vol. 23, no. 23, pp. 12491–12510, Dec. 2019. doi: 10.1007/s00500-019-03794-x
[54]	A. Kumar, G. Wu, M. Z. Ali, Q. Luo, R. Mallipeddi, P. N. Suganthan, and S. Das, “A benchmark-suite of real-world constrained multi-objective optimization problems and some baseline results,” Swarm Evol. Comput., vol. 67, Art. no. 100961, Dec. 2021. doi: 10.1016/j.swevo.2021.100961
[55]	Y. Tian, Y. Zhang, Y. Su, X. Zhang, K. C. Tan, and Y. Jin, “Balancing objective optimization and constraint satisfaction in constrained evolutionary multiobjective optimization,” IEEE Trans. Cybern., vol. 52, no. 9, pp. 9559–9572, Sep. 2022. doi: 10.1109/TCYB.2020.3021138
[56]	Y. Tian, T. Zhang, J. Xiao, X. Zhang, and Y. Jin, “A coevolutionary framework for constrained multiobjective optimization problems,” IEEE Trans. Evol. Comput., vol. 25, no. 1, pp. 102–116, Feb. 2021. doi: 10.1109/TEVC.2020.3004012
[57]	Y. Tian, R. Cheng, X. Zhang, and Y. Jin, “PlatEMO: A MATLAB platform for evolutionary multi-objective optimization[educational forum],” IEEE Comput. Intell. Mag., vol. 12, no. 4, pp. 73–87, Nov. 2017. doi: 10.1109/MCI.2017.2742868
[58]	P. A. N. Bosman and D. Thierens, “The balance between proximity and diversity in multiobjective evolutionary algorithms,” IEEE Trans. Evol. Comput., vol. 7, no. 2, pp. 174–188, Apr. 2003. doi: 10.1109/TEVC.2003.810761
[59]	L. While, P. Hingston, L. Barone, and S. Huband, “A faster algorithm for calculating hypervolume,” IEEE Trans. Evol. Comput., vol. 10, no. 1, pp. 29–38, Feb. 2006. doi: 10.1109/TEVC.2005.851275