Flow fluctuation suppression of mud lifting pump based on fractional-order sliding mode control

In deep-sea riserless mud recovery (RMR) drilling operations, single-pump mud lifting systems lack sufficient equipment redundancy under deepwater environmental conditions and harsh operating scenarios. They also demonstrate limited capability in handling complex downhole incidents such as lost circulation and kicks. While dual-pump collaborative operations enhance the adaptability to complex deepwater conditions, flow rate fluctuations during pump switching processes may induce cuttings deposition, posing severe operational stability challenges. To address these issues, first, a discrete element fluid-solid model for deep-sea lifting pumps is developed through coupled Fluent and EDEM (Engineering Discrete Element Method) simulations, deriving time-dependent flow rate characteristic curves during pump startup; second, a multi-scenario smooth flow control model for dual-pump systems is designed based on a closed-loop dual-pump configuration, enabling numerical experimentation on pipeline flow variations under diverse operational conditions, rotational speed, and valve actuation parameters. Simulation analysis examines the relationship between actual monitored flow rate and preset target value; third, a fractional-order sliding mode control law is designed based on flow error signals, mapping these errors to parameter adjustments for pump flow regulation. By optimizing control parameters and strategies, system smoothness and stability during operational mode switching are significantly improved. Finally, numerical validation demonstrates the efficacy of the proposed strategy. In typical scenarios of switching from single pump operation to dual-pump series operation and from to dual-pump parallel operation, compared with PID controller models and sliding mode control (SMC) respectively, the model based on fractional-order sliding mode control (FOSMC) significantly reduces the flow deviation, remarkably reducing flow fluctuations. Results show the FOSMC strategy achieves faster convergence speed and lower steady-state errors. This study develops a fluid-solid coupling prediction model to systematically analyze flow characteristics under various operating conditions through systematic prediction and thoroughly understand flow fluctuations. Building on this, the proposed fractional-order control framework not only enhances flow stabilization compared to traditional methods but also optimizes optimal flow control strategies. The results achieve dynamic flow stability in the RMR system and support the design of a high-performance, reliable deep-sea mud lifting pump. Additionally, they provide a theoretical foundation for subsequent experiments-particularly the transient behaviors during pump switching, while offering a validated control framework for the dual-pump mud lifting system.