An analysis method based on pipeline state data is proposed

which can get the pipeline deterioration trend and help to prevent the leakage caused by pipeline deterioration. The specific method is to quantitatively construct the health model of the improved Mahalanobis distance of the primary pipeline by integrating the advantages of the analytic hierarchy process of subjective expert experience and the entropy weight method of objective big data analysis. Then

the combined prediction model of convolution neural network and long-term and short-term memory neural network is applied to analyze and predict the future deterioration and leakage of the primary pipeline. The simulation results show that

the health model constructed in this paper can better reflect the real health state of the primary pipeline. The combined prediction model can effectively predict the trend of deterioration and leakage of primary pipeline

which provides a reference for ensuring the safe and stable operation of equipment.