| 摘要 |
<p>An adaptive method for detecting the heartbeat period of messages, the method comprising: a construction step for constructing a neural network model, input layer nodes thereof being heartbeat period parameters influencing the heartbeat period and output layer nodes thereof being the heartbeat period, and the sample set of the neural network model including the mapping relationship between the sample parameters of the heartbeat period and the sample value of the heartbeat period; a self-learning triggering step for triggering the neural network model to perform self-learning according to the sample set when it is determined that the triggering conditions of self-learning are satisfied; a calculation triggering step for collecting parameters for calculating the heartbeat period and inputting them into the neural network model when it is determined that the triggering conditions of heartbeat period calculation are satisfied, so as to obtain a calculated value of the heartbeat period; a heartbeat period updating step for updating the current heartbeat period according to the calculated value of heartbeat period when it is determined that the triggering conditions of heartbeat period update are satisfied, or else performing a sample set updating step; and a sample set updating step for updating the sample set, and then returning to the self-learning step.</p> |
