Power battery production process fluctuation anomaly detection method based on space-time diagram model

The invention provides a power battery production process fluctuation anomaly detection method based on a space-time diagram model. The method comprises the steps that key parameters of a battery production process are used for model training, and time convolution networks with different convolution kernel sizes are used for obtaining a time model of a sequence; constructing a leading matrix, and obtaining spatial information of the time sequence by using a graph convolutional neural network; constructing a space-time diagram convolution block with a gating mechanism to filter the obtained information to obtain effective time and space dependence; aggregating output information of all the gated space-time diagram convolutional networks to perform single-step prediction on the input sliding time window; calculating a prediction error by using the observation value and the prediction value of the data, and then calculating a threshold value of an abnormal score by using the prediction error; and if a prediction