Bridge health early warning method based on time sequence and multi-sensor fusion

The invention relates to a bridge health early warning method based on time series and multi-sensor fusion. Selecting W sensor data streams of a public bridge as a data set; checking the data to determine that the data set is an available data set; building an ARMA model and carrying out order determination on the model; performing data processing on the available data set to obtain a new data set; removing abnormal sensor data in the new data set to obtain ARMA model input data; initializing a fixed-order ARMA model and training the model by adopting long-time monitoring data to obtain a pre-trained ARMA model; presetting an alarm level, and performing step-by-step updating on the pre-training model by using the short-time monitoring data; and taking all monitoring data of the target bridge at the current time as model input after step-by-step updating, outputting and obtaining prediction results of various types of monitoring indexes of the target bridge, then carrying out information fusion on the predicti