Journal of Engineering Research and Reports

Cement powder weighing accuracy in concrete mixing stations directly impacts concrete performance and quality, with airborne residual material-induced errors being the key constraint on accuracy improvement. This paper proposes an iterative learning control-based cement weighing optimization strategy to compensate for airborne residual errors by dynamically adjusting feed closing advance. Firstly, it analyzes the dynamic characteristics of the batching weighing system, clarifying the formation mechanism of airborne residual material and its impact on weighing accuracy. Then, it designs the core optimization law of iterative learning control: through real-time error feedback and historical data learning, the strategy autonomously optimizes high-speed switching and stop advance parameters, with learning and forgetting factors selected to balance convergence and stability. Taking fixed threshold control and feedforward PID as comparison schemes, theoretical analysis is carried out from parameter adaptation, model dependence and computational complexity. MATLAB simulation results show that the strategy can converge stably with fewer iterations under multi-objective weight conditions, the error meets the industrial high-precision requirements, and the efficiency is equivalent to the traditional method. Its core innovation is to get rid of the dependence on accurate modeling, and to achieve robust adaptive control under multi-objective conditions through iterative self-learning, which provides an efficient solution for improving the batching accuracy of the mixing station, and has reference value for similar batch weighing systems.