Operational Phase Recognition and Power Matching under Shoveling Conditions in an Autonomous Loading Mobile Concrete Mixer

As a representative working condition for loading-type construction machinery, the shoveling operation is characterized by highly variable operational demands and rapidly changing loads, which significantly complicate power matching across different phases of the process. This study proposes a power matching strategy based on phase-specific recognition and dynamic torque regulation, grounded in the identification of the shoveling stages of an automated loading mobile concrete mixer and the analysis of its load characteristics under shoveling conditions. The distinct phases of the shoveling process were correlated with system pressure and power features, and an intelligent recognition model was established using a bidirectional long short-term memory neural network algorithm. By analyzing the engine’s power performance, fuel economy, and the efficiency of the variable displacement pump, a constrained multiobjective optimization equation was formulated to balance both the dynamic performance and energy efficiency of the automated loading mobile concrete mixer. This approach reduces the absolute torque deviation across different operational stages to within 50 Nm. Peak fuel consumption decreased to 2.5 g/s. The proposed strategy effectively addresses the mismatch in power demands across varying operational phases and simultaneously reduces overall system energy consumption.