Rapid transient operation control method of natural gas pipeline networks based on user demand prediction

The natural gas pipeline networks play a vital role in Integrated Energy System (IES). Simultaneously, with the increase in the number and types of users, the operational model needs to be transformed from plan-oriented to demand-oriented. Hence, higher requirements are put forward for the flexibility and rapidity of the control methods. In this paper, a novel rapid operation control method for natural gas pipeline networks is proposed based on multi-user demand prediction and pipeline flow state inversion. First, the Nonlinear autoregressive (NAR) neural network is used to analyze the historical consumption of various users and predict each user's demand in the short term. Second, the rapid transient inversion method is derived from the partial differential control equation to realize the inverse of the pipeline flow state. Third, combined with the equipment model, the control method is formed to realize rapid control of the pipeline networks. The control method is applied to classical triangular networks and existing pipeline networks to demonstrate accuracy and effectiveness. The time-consuming ratio of the commercial software and proposed model for the triangular pipeline networks is 1:0.22. For the existing pipeline networks, six users' average demand prediction errors are 5.58%, and the control errors of pressure at three markets are 1.00%, 1.22%, and 0.81%. The results show that the proposed method can respond to users' demands rapidly and provide decision support for the safe and efficient operation of the natural gas pipeline networks.

Automated summary

A novel rapid operation control method for natural gas pipeline networks is proposed based on multi-user demand prediction and pipeline flow state inversion. The method analyzes the historical consumption of various users and predicts their demand in the short term using a Nonlinear autoregressive (NAR) neural network. The rapid transient inversion method is derived from the partial differential control equation to realize the inverse of the pipeline flow state, and combined with the equipment model to achieve rapid control of the pipeline networks.