Learning-Based Model Predictive Control (LMPC) is a class of algorithms that enhances Model Predictive Control (MPC) by including machine learning methods, improving robot navigation in complex environments.However, the combination of machine learning and MPC computation in LMPC creates a unique workload that cannot be efficiently handled by a simple GPU and CPU integration.We present HiPER, a hierarchically-composed processing array that provides temporal and spatial mapping capabilities, allowing efficient adaptation to workload changes at runtime.To simplify control, HiPER employs a pointer queue hierarchy to compose and orchestrate program execution.Additionally, HiPER utilizes a fractal interconnect topology that combines local systolic interconnects and their hierarchical extensions to efficiently support the workload's traffic characteristics.To evaluate the performance and efficiency of HiPER, we synthesized a 16.37 mm 2 design in 16nm CMOS.The design consists of 6 pointer queue levels and 1024 PEs.The prototype was assessed using a representative LMPC workload, demonstrating 10.75× improvement in performance compared to a GTX 1080 GPU, 12.80× improvement in energy efficiency compared to a Jetson Orin Nano embedded GPU, and 11.6×/22.2×improvement in performance compared to the RoboX/Plasticine accelerators.

Accurate truck arrival prediction is complex but critical for container terminals. A deep learning model combining Gated Recurrent Unit (GRU) and Fully Connected Neural Network (FCNN), is proposed to predict daily truck arrivals using fusion technology. The model can efficiently analyze sequence and cross-section data sets. The new feature in the new model lies in that it, for the first time, incorporates the new parameters influencing traffic volumes such as the vessel-related information, arrival weekdays, and weather conditions into the long-time series of truck arrivals. Furthermore, truck arrivals are predicted in three groups based on their movement purposes: pick-up, delivery, and dual. it also contributes to the literature in a sense that the performance of the model is tested using real big data from a world-leading container port in Southern China. The results generate insightful managerial implications for guiding port traffic management in a generic manner. It reveals the relation of export container arrivals with the Container Yard (CY) closing time of a specific vessel. It is demonstrated the proposed model outperforms the currently available methods with an improved accuracy rate of prediction by 23.44% (dual), 32.09% (pick-up), and 26.99% (delivery), respectively. As a result, the model can better reflect reality compared to the existing ones in the literature. It is also evident that the 3-categorized prediction model can significantly help increase prediction accuracy in comparison with the 2-categorized methods used in practice.