Sensor data classification method based on Stack integrated neural network

The invention discloses a sensor data classification method based on a Stack integrated neural network, and the method comprises the steps: 1) constructing a basic learner which comprises a feature extraction unit and a feature classification unit connected with the feature extraction unit, 2) taking outputs of a plurality of basic learners as inputs of meta learners, and constructing a Stacking model, and 3) simultaneously inputting the sensor data into a one-dimensional convolution layer of each feature extraction unit of each basic learner, taking the output of each basic learner as the input of a meta-learner of the Stacking model, processing the sensor data by the meta-learner, and outputting a final classification result by a softmax function layer of the meta-learner. According to the method, the advantages of the convolutional neural network and the long-term and short-term memory network for extracting the sensor data features are combined, the accuracy of sensor data classification is improved, and