Intelligent control methods for autonomous transportation platforms based on long short-term memory networks (LSTM)

With the rapid development of industrial logistics and intelligent transportation systems, unmanned transportation platforms are increasingly widely used in complex dynamic environments due to their high efficiency, flexibility and intelligence. However, traditional control methods (such as PID control and model predictive control) show problems of insufficient adaptability and low prediction accuracy when facing nonlinear and dynamically changing scenarios. To address this challenge, this paper proposes an intelligent control method for unmanned transportation platforms based on long short-term memory (LSTM) networks. By constructing a time series prediction model, high-precision prediction and path planning of the platform's dynamic behavior are achieved. Experimental results show that the LSTM model is significantly superior to traditional methods and RNN in three typical scenarios (static environment, dynamic obstacle environment and high noise environment). In the dynamic behavior prediction task, the mean square error (MSE) of the LSTM model is reduced by at least 70% compared with the traditional method, and the success rate in the path planning task is increased by 15%-36%. In addition, the LSTM model shows strong robustness in a high noise environment and effectively copes with complex and changing dynamic scenarios.