The growing diversity and number of industrial robots make energy consumption prediction and optimization increasingly essential. Current data-driven approaches, particularly those based on multi-layer perception (MLP), have shown feasibility but typically overlook the variability or unknown nature of load-related parameters in real-world applications. This paper presents a KAN-LSTM model designed to accurately predict energy consumption under unknown load conditions, alongside a particle swarm optimization (PSO) algorithm for minimizing energy use. First, an industrial robot dynamics and energy consumption model is established. Then, the KAN-LSTM model is trained on datasets from the AUBO-E5 robot, with its predictions compared to alternative network models. Finally, PSO is applied to optimize energy consumption. Experimental results indicate that the KAN-LSTM model achieves high prediction accuracy (95.7–97.1%) and offers substantial energy optimization potential (53.1–64.7%). Optimized industrial robots are particularly suitable for tasks such as picking and palletizing in the courier industry, saving operational costs and increasing the sustainability of automated systems in logistics environments.

Given the continual rise in mission diversity and environmental complexity, the adept integration of a robot’s aerial and terrestrial locomotion modes to address diverse application scenarios has evolved into a formidable challenge. In this paper, we design a reconfigurable airframe robot endowed with the dual functionalities of rolling and flying. This innovative design not only ensures a lightweight structure but also incorporates morphing capabilities facilitated by a slider-crank mechanism. Subsequently, a land-to-air transformation strategy for the robot is introduced, achieved through the coordinated movement of the robotic arm and the servo motor. To ensure stable control of the robot amid external wind disturbances, we leverage the collaboration between a Generative Adversarial Network (GAN)and a Nonlinear Model Predictive Control (NMPC) controller. After the wind force magnitude is predicted through the neural network, the robot’s adeptness in flexible trajectory tracking is verified. Under simulated wind conditions of 12.1 m/s, the trajectory error consistently remains within the range of 10–15 cm, affirming the effectiveness of this control method.