This paper presents a unique Modified Differential Evolution Optimization Algorithm (MDEOA) for the intelligent modification of proportional–integral–derivative (PID) controller settings. The proposed MDEOA is developed with particular emphasis on future PID controller tuning applications. Benchmark evaluations demonstrate significant improvements in convergence speed and optimization quality compared with Classical Differential Evolution (CDE), indicating its potential suitability for intelligent control parameter optimization. The study tackles the enduring drawbacks of CDE in control parameter optimization, including its inadequate exploration-exploitation balance and sluggish convergence when adjusting dynamic or nonlinear systems. The suggested MDEOA addresses these issues by introducing improved crossover, mutation, and adaptive control techniques that increase population diversity and hasten convergence toward ideal PID gains. To give a methodical comparison analysis, the PID tuning problem is subjected to both the traditional CDE and the suggested MDEOA. The MDEOA-based PID controller offers noticeably better dynamic performance, as shown by analytical simulations and experimental validation. In particular, it completely eliminates the maximum peak and steady-state error, corresponding to 100% reduction in both parameters. The MDEOA-based PID controller reduces the rise time by 10.18%, the peak time by 49.70%, and the settling time by 93.47% compared with the conventional PID controller. These enhancements are a direct result of the algorithm's modified operators' efficacy. Overall, the findings show that MDEOA is a strong and dependable substitute for conventional CDE in intelligent control system parameter optimization. MDEOA is a viable tool for future applications in sophisticated automation and real-time control contexts because of the large improvements in response characteristics, which demonstrate that the suggested alterations greatly increase the resilience, accuracy, and flexibility of PID tuning.

Engineered cementitious composites (ECCs) are increasingly considered for protective structures because their fiber-bridging mechanism, tensile ductility, and energy absorption can restrict crack growth under severe loading. Their compressive response at high strain rates, however, remains difficult to represent using conventional concrete constitutive models, particularly when rate-dependent deformation and progressive damage occur concurrently. This study investigated the dynamic compressive behavior of ECC containing 2.0% polyvinyl alcohol fibers and developed a constitutive framework that couples nonlinear viscoelasticity with statistical damage evolution. Split Hopkinson pressure bar (SHPB) tests were conducted at four average strain-rate levels ranging from approximately 15 to 200 s$^{-1}$, with three specimens tested at each level. Relative to the lowest strain-rate level, the dynamic peak stress and peak strain increased by up to 41.0% and 98.6%, respectively, while their rates of increase gradually declined at the higher loading rates. A rate-dependent constitutive model was then formulated by combining the Zhu–Wang–Tang (ZWT) nonlinear viscoelastic model with a Weibull damage function. The model reproduced the ascending branch and peak region of the measured stress–strain curves, although larger discrepancies remained in the post-peak softening stage. After implementation in Livermore Software for DYnamic Analysis (LS-DYNA), the model reproduced the strain-rate-dependent stress–strain response and the transition from localized cracking to extensive fragmentation. The maximum deviations between the simulated and experimental increases in peak stress and peak strain were 8.2% and 6.5%, respectively. The proposed framework provides a physically interpretable representation of the coupled rate-dependent deformation and damage of ECC and supports numerical analysis of ECC components subjected to impact-type loading.