Zhengyi Wang, Xinsheng Wang, Miao Ren
Crushing discarded brick to create recycled brick aggregates (RBAs) has emerged as a practical method for mitigating environmental degradation and solving the scarcity of natural resources in civil engineering. This research aims to encourage the extensive use of recycled brick aggregate concrete (RBAC) in the building industry. To do this, the elastic modulus of RBAC (ERBA) is estimated using coupled fuzzy networks. The elastic modulus of recycled brick aggregate concrete is determined by six input variables. These variables include the standard compressive strength of cement paste (fcem), the mass-weighted water absorption ratio of coarse aggregates (ωmwa), the effective water-to-cement ratio of the fresh mixture (weff/ c), the sand-to-aggregate ratio (s/a), the volume replacement ratio of recycled brick aggregates (ηRBA), and the cylindrical compressive strength (fcy). These variables are identified using a computational database that contains 123 test results from previous studies. To be more exact, the procedure of developing and assessing the proposed framework included utilizing 75% of the data as a training set and the remaining 25% as a validation set. This work conducts a comparison between two hybrid optimum methodologies, namely the Coati algorithm (CA) and the Chaos game algorithm (CGA), when combined with a traditional machine learning methodology called the Adaptive neuro-fuzzy inference system (ANFIS). For the learning and assessment stages, the CAA showed U95 index values of 4.157 and 3.0541, correspondingly. For the CGAA, the corresponding U95 values were 4.6391 and 3.3734. These conditions remained throughout the whole project. This illustrates that the CAA has the capability to precisely predict both proficiency and dependability. Both models are very precise and reliable, with the CAA model being somewhat better, as shown by logical reasoning and evaluation indications. Received: 20.01.2025 Received in revised form: 19.05.2026 Accepted: 14.07.2026