Evaluation of AS/RS based on evidence theory and BP neural network

In order to improve the efficiency of AS/RS system, the evaluation index system of AS/RS system is established from the system of economy, stability, technicality and speed, sociality and safety. Based on the index system, evaluation model is established for AS/RS system combining the BP neural network and evidence theory. In this method, the normalized warehouse data as input of BP neural network, through the different parameters of the neural network training the output value has been obtain. Using D-S evidence theory synthesis rules for the output value, evaluation result has been obtained. The method solves the output instability of a single BP neural network and overcame the evidence theory subjectivity through the expert knowledge to determine the evidence, improved the credibility of the evaluation results. Using the evaluation model do an empirical analysis of four AS/RS, the result has shown the feasibility and effectiveness of the model in the AS/RS system.