Periodic Fractional Control in Bioprocesses for Clean Water and Ecosystem Health

Abstract: This paper presents a novel fractional-order chemostat model (FOCM) for optimizing biological water treatment processes, incorporating memory effects through the use of Caputo fractional derivative with sliding memory (CFDS). Traditional integer-order models fail to capture the time-dependent behaviors and memory effects inherent in microbial systems. Our work addresses this limitation by developing a fractional-order framework that represents microbial growth and substrate degradation dynamics more accurately. The primary objective is to minimize the average pollutant concentration in treated water through optimal periodic control (OPC) of the dilution rate, subject to constraints on treatment capacity and periodic boundary conditions. Key contributions include: (1) reduction of the 2D fractional-order system to a computationally tractable 1D fractional differential equation while preserving essential dynamics; (2) rigorous proof of the existence and uniqueness of optimal periodic solutions using Schauder's fixed-point theorem and convexity arguments; (3) derivation of bang-bang optimal control (OC) strategies by using a fractional Pontryagin maximum principle (PMP); and (4) comprehensive numerical simulations demonstrating significant performance improvements over steady-state operation. Our results show that periodic fractional control can reduce average substrate concentrations by up to 40% compared to steady-state operation, with the fractional order {\alpha}, the dynamic scaling parameter $\vartheta$, and the sliding memory length L serving as critical factors that govern memory effects, control responsiveness, and switching frequency. The proposed framework bridges fractional calculus with environmental engineering, offering new insights for designing sustainable water treatment systems with improved pollutant removal efficiency.