Layot Optimizaioin Method Research Of Natural Gas Network Based On The Minimum Risk Loss Cost

Urban natural gas?NG?pipeline network system is one of the main infrastructure of the city,since NG Has gradually replaced conventional energy sources,such as oil and gas,and became the main type of energy supply in cities.Urban natural gas networks have flammable and explosive properties and are usually laid in central urban areas where people are more concentrated in areas with important public buildings and infrastructure.Thus the leakage of natural gas networks will cause huge direct and indirect economic losses,including personnel public property safety and environmental pollution,etc,that caused by fire or explosion accidents.At present,the layout planning of urban natural gas pipeline network in actual engineering is mainly based on the pipeline laying route,that can be laid and determined by the subjective consciousness of the designer.In the theoretical research,there are related literatures on the optimization of branch pipe network layout and the annular pipe network optimizationand,for which the optimization objective is shortest pipe length and supply reliability.However,there is little research on layout optimization for minimizing risk loss during the layout planning phase.In addition,there is no risk evaluation theory for the layout planning stage of the proposed pipeline network,for all the risk assessment methods are based on the risk assessment of the established pipeline network.A layout optimization method based on minimizing risk loss and its corresponding verification method are proposed in this paper,in order to minimize the risk loss during the planning phase.Firstly,The fault tree analysis?FTA?for the failure factor of urban natural gas network is modeled,annd then 92 failure factors are determined as input variables of training network based on practical experience,and the theoretical failure probability value is calculated based on FTA model,which is used as output variable.BP?Back Propagation Neural Network?and RBF?Radial Basis Function Neural Network?prediction models for failure probability?BP-FPM and RBF-FPM?,which have the best combination of network parameters,are finally determined by trial and error.The two prediction models are applied to the same example.The two prediction results and the traditional failure probability results,which is calculated by employing FTA,are comared to confirme the RBF prediction model with the smallest error is the failure probability prediction model?RBF-FPM?.Then,the internal correspondence between soil corrosion grade and economic zoning of failure consequence has been found through repeated tests,based on the two partition standard,so that the fuzzy calculation model of failure consequence is put forward.The optimal independent variable combination scheme was determined by analyzing the relationship between soil composition and economic loss of failure consequences based on correlation analysis technique.The fuzzy prediction models of the failure?FPF?consequence are compared by using the three prediction techniques,i.e.,fitting,regression and neural network.Finally,the three fuzzy prediction models with the highest accuracy are determined as follows:y=259.156+991.215₁+36.331₂-0.829₃?y=444.61815-911.97361x+399.86108x2+138.8713x3-81.96061x4 and BP neural network model of training function trainlm,the corresponding decision coefficient R2 is:0.864?0.890 and 0.944.For an engineering example,the soil corrosion grade is calculated based on two method,one is entropy weight method,and the other is the three selected FPF and FCM.Through the comparison and analysis of the soil corrosion grade determined by the traditional method and the new model,BP-FPF with the highest precision is determined,and the correctness and feasibility of FCM are verified.The fuzzy prediction model of risk loss?FPR?is put forward,aased on the definition of risk loss cost and combined with the two selected RBF-FPM and BP-FPF).Then,The risk loss fuzzy prediction value?RFV?,which is calculated by FPR,is used to design the equivalent cost lengthl,that could minimizes the risk loss in the layout planning stage.Thus,the model of branch pipe network layout optimization and loop pipe network layout optimization can be established,which can minimize the risk loss in the layout planning stage,by improving the equivalent length based on the FPR and the distribution optimization physical model.By comparing and analyzing various spanning tree algorithms in graph theory,the Kruskal algorithm of minimum spanning tree is selected as the algorithm to solve the optimal mathematical model of branch pipe network layout.Two intelligent algorithms,Ant Colony Optimization?ACO?and genetic algorithm?GA?,are applied to solve the two difficult problems in the optimization of loop network layout.Finally,the optimal layout of the two networks with minimum risk loss fuzzy prediction value?RFV?is determined by using the prepared program.Then,the verification method of the layout optimization is established based on the traditional contrast verification method and the characteristics of the layout optimization.Moreover,the optimum layout?LT?of minimum traditional risk loss cost?TRC?is determined by employing the theoretical foundation of the natural gas pipeline network parameter optimization and its corresponding GA.First of all,two kinds of layouts,RFV minimum layout?L1?and ethe tshortest path layout?L2?,which are determined by traditional theory,are determined by using the layout optimization method proposed in this paper for an intermediate pressure ring pipe network.And then the construction costs,TRC and and the combined cost are calculated and the percentage of the difference between L1 and L2 are 7.55%and 7.48%respectively,based on traditional calculation method of risk loss cost?TMR?and GA.Finally,it is concluded that the optimization results of the loop network parameters are different from those of the branch network,through the comparative analysis of the three kinds of cost.This is because the objective function of the loop layout optimization not only refers to the shortest path,but also needs to satisfy the best reliability.The results of the loop network example show that the synchronous optimal layout in this paper can not only minimize the risk loss,but also reduce the construction cost and the corresponding comprehensive cost when applied to loop network.The layout optimization based on risk loss minimization has obvious economic benefits.Finally,the synchronous optimization method,especially the core technology,is demonstrated through an intermediate pressure branch pipe network system.The core technology includes the following key steps:failure probability prediction,risk loss fuzzy prediction,solution of two kinds of optimal layout.Parameter optimization and traditional risk loss calculation of two optimal layouts.The TRC of L1 is 11.709%less than that of L2,and the total cost of L1 is 13.917%more than that of L2.By comparing the sum of the two kinds of cost?the comprehensive cost?for the two layouts,it is shown that although the minimum risk loss of this instance is the least,and the construction cost is not the least,the comprehensive cost of the former is about 2.308%less than the latter.The layout optimization method of risk loss minimization,which is established in this paper,can be used to minimize the risk loss at planning phase for natural gas network.Through the comparative analysis of three costs for two layouts,the decision-makers can better weigh the advantages and disadvantages between the risk cost and the construction operation cost,and put forward a more targeted and economic optimal layout scheme.At the same time,feasible solutions and theoretical support for the study of risk loss are provided in the planning and design stage of other fields.