The Offset Azimuth Prediction of Light Buoy Based on Markov Chain Optimization Multiplicative Seasonal Model

Aimed at the problem of large error caused by uncertain factors in fitting process traditional multiplicative seasonal model, advantages Markov chain this study are applied to model optimize prediction results. Based on residual value between theoretical and actual values, values different intervals divided into states. The transition probability matrix is established through probabilities; then, weighted sum probabilities carried out select optimal state. real number Meizhou Bay portlight buoys used verify effect MAE, MAPE, RMSE, RRMSE, SSE, R2 calculate predicted value. results show that compared other models, MAE MC-SARIMA decreased 2.19003794, MAPE 0.66%, RMSE 2.092671823, RRMSE 0.006221352, SSE 404.0231931, increased 0.224686247. It shows optimized can predict azimuth data light buoy more effectively than models.