Pipe network compressor station merging and simplifying operation optimization method based on response surface method

Abstract

The invention discloses a response surface method-based pipe network compressor station merging and simplified operation optimization method, which comprises the following steps of: constructing a global pipe network topological structure, and identifying mergeable local pipe sections of multiple compressor stations; collecting basic operation data of a pipe network, and determining an operation condition range of a local pipe section; the method comprises the following steps: carrying out preliminary simplification merging treatment on mergeable local pipe sections of multiple compressor stations, and constructing a multi-working-condition optimal operation scheme set and a response model of the local pipe sections by adopting a response surface experimental design and analysis method; the global pipe network topological structure and the optimization model are combined and simplified, and an optimal operation scheme of the global pipe network is solved; adopting an approximate working condition selection method to select the optimal operation scheme of the local pipe section under the current working condition; and quickly solving an optimal operation scheme according to the latest operation condition. According to the method, the optimization solving efficiency of pipe network operation can be effectively improved, the operation condition range is fully covered by combining response surface experiment design and analysis methods, and rapid and reliable decision support is provided for real-time operation scheduling of the pipe network.

Description

Technical Field

[0001] The present invention relates to the field of operation optimization of natural gas pipelines, and particularly to an operation optimization method for merging and simplifying pipeline compressor stations based on the response surface method. Background Art

[0002] Natural gas pipelines are important infrastructure for energy transportation. Their operation optimization technology plays a key role in ensuring energy supply security and improving energy utilization efficiency. At the same time, it is also of great significance for reducing the operation cost of pipelines and achieving low-carbon environmental protection goals. With the continuous growth of natural gas demand and the continuous expansion of pipeline network scale, the coordinated operation control of a large number of compressor stations along the line has gradually become the core issue in this field of research. Compressor stations are not only key facilities to ensure the stable transportation of natural gas, but also the main sources of energy consumption and carbon emissions in pipeline network operation. The rationality of their operation plans directly affects the economy and safety of the pipeline network system. However, with the increase in pipeline network scale and the complexity of the structure, the solution dimension and complexity of the pipeline network operation optimization problem have also increased significantly. In addition, due to the influence of multiple factors such as fluctuations in gas source production and changes in user demand, the operating conditions of natural gas pipelines are dynamically variable, and it is necessary to quickly adjust the operation plan in real time to adapt to the changes in operating conditions. This poses higher requirements for the solution efficiency and timeliness of pipeline network operation optimization methods, and there is an urgent need to establish an efficient and reliable optimization method to cope with the complex and variable operating environment of natural gas pipelines.

[0003] At present, a large number of scholars have carried out relevant research in the field of operation optimization technology of natural gas pipelines. By establishing optimization models and combining traditional optimization algorithms such as dynamic programming method and heuristic stochastic optimization algorithms such as genetic algorithm for solution, a variety of pipeline network operation optimization methods have been formed. Relevant applications show that these methods can optimize the pipeline network operation plan to a certain extent, and are mainly applicable to small-scale natural gas pipelines with simple structures and low solution dimensions, and can quickly solve and effectively reduce the operation cost of the pipeline network. However, when applied to large-scale natural gas pipelines that are increasingly complex and large-scale, the existing optimization methods are difficult to find a feasible and relatively optimal operation plan in the high-dimensional optimization space, and the solution time is too long, making the existing methods unable to meet the application requirements of the actual operating conditions of the pipeline network and the timeliness requirements of operation scheduling, resulting in ineffective application in the actual operation scheduling of the pipeline network.

[0004] In view of this, a method for optimizing the combined and simplified operation of pipeline network compressor stations based on the response surface method is established. By effectively identifying the locally combinable pipe segments of multiple compressor stations in the natural gas pipeline network, combining the response surface experimental design and analysis methods, constructing the optimal operation plan set for multiple working conditions of the locally combinable pipe segments, fitting the response model between the working condition parameters and operation costs of the locally combinable pipe segments, and then respectively performing combined and simplified processing on the global pipeline network topology structure and the optimization model, the dimensionality reduction and simplification of the operation optimization problem of large-scale natural gas pipeline networks are realized, which can significantly reduce the optimization space and calculation burden of the operation optimization problem of large-scale pipeline networks, greatly improve the solution speed, effectively support quickly obtaining the optimal operation plan for real-time operation conditions, and meet the timeliness requirements of pipeline network operation scheduling. Summary of the Invention

[0005] In view of the above problems, the purpose of the present invention is to provide a method for optimizing the combined and simplified operation of pipeline network compressor stations based on the response surface method, which can realize quickly and reliably solving the optimal operation plan with the lowest operation cost under real-time conditions through the dimensionality reduction and simplification processing of the global pipeline network operation optimization problem.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A method for optimizing the combined and simplified operation of pipeline network compressor stations based on the response surface method, which includes the following steps:

[0007] Step 1: Collect the basic data of the pipeline network structure, construct the global pipeline network topology structure, and use the identification method for simplifiable and combinable compressor stations to identify the global pipeline network topology structure to obtain the locally combinable pipe segments of multiple compressor stations;

[0008] Step 2: Collect the basic operation data of the pipeline network and determine the operation condition range of each locally combinable pipe segment of multiple compressor stations;

[0009] Step 3: Traverse and analyze each locally combinable pipe segment of multiple compressor stations. If there are branch pipelines or two or more distribution stations or uploading stations along the locally combinable pipe segment, according to the distribution of station locations and the operation condition range, use the preliminary simplification and combination processing method to simplify the branch pipelines and approximately combine multiple distribution stations or uploading stations to obtain the locally combinable pipe segments of multiple compressor stations after preliminary simplification and combination;

[0010] Step 4: For each locally combinable pipe segment of multiple compressor stations, combine the operation condition parameters and range to establish the corresponding factor level table, and use the Box-Behnken method to design the response surface experimental plan including multiple groups of different operation condition parameter combinations;

[0011] Step 5: Using the optimization model for minimizing the pipeline network operation cost, for each locally mergeable pipe section of multiple compressor stations, solve the optimal operation plan and the minimum operation cost for each combination of operation condition parameters in the response surface experiment plan, construct a set of optimal operation plans for multiple working conditions of the local pipe section, and fit the response model between the working condition parameters and the operation cost of the local pipe section;

[0012] Step 6: Using the optimization model for minimizing the pipeline network operation cost, solve the global pipeline network operation optimization problem, and using the response model established in Step 5, perform merging and simplification processing on each locally mergeable pipe section of multiple compressor stations in the global pipeline network topology and the optimization model, and combine with the optimization algorithm to solve and obtain the optimal operation plan of the global pipeline network;

[0013] Step 7: According to the optimal operation plan of the global pipeline network in Step 6, determine the current operation condition parameters of each locally mergeable pipe section of multiple compressor stations, and using the approximate working condition selection method, correspondingly select the most approximate plan from the set of optimal operation plans for multiple working conditions of the local pipe section as the optimal operation plan of the current local pipe section;

[0014] Step 8: After completing Steps 1 - 5, when the operation condition of the pipeline network changes, according to the latest operation condition parameters, re - perform Steps 6 and 7, and then the optimal operation plan with the lowest operation cost under the real - time working condition can be quickly solved and obtained.

[0015] Furthermore, in Step 1, the method for identifying mergeable compressor stations includes the following steps:

[0016] S11: According to the global pipeline network topology, the pipeline flow direction, and the distribution of compressor stations, identify the pipeline intersection nodes. If there is more than one pipeline along the pipelines connected by the intersection node with a compressor station, split the global pipeline network topology at this pipeline intersection node, and traverse all pipeline intersection nodes to obtain the local pipeline topologies;

[0017] S12: Traverse and analyze the local pipeline topologies obtained in S11. If the local pipeline topology contains two or more compressor stations, starting from the first compressor station node at the upstream of the local pipeline and ending with the last compressor station at the downstream of the local pipeline, further delimit to obtain the locally mergeable pipe section of multiple compressor stations.

[0018] Furthermore, in Step 2, the operation condition range of the locally mergeable pipe section of multiple compressor stations includes: the inlet pressure operation range of the starting compressor station, the outlet flow operation range of the ending compressor station, and the outlet pressure operation range of the ending compressor station, as well as the download volume operation range of the distribution stations along the local pipe section and the upload volume operation range of the upload stations.

[0019] Furthermore, in Step 3, the preliminary simplification and merging processing method includes the following steps:

[0020] S31: If there are branch pipelines along the local pipeline section, the operating range of the download volume of the sub - stations along the branch pipelines and the operating range of the upload volume of the upload stations are respectively accumulated. If the total download volume is larger, all the stations along the branch pipelines are merged into one sub - station in the local pipeline section topology; conversely, if the total upload volume is larger, they are merged into one upload station. The merged station is set at the connection of the branch pipeline and the local pipeline section, and the absolute value of the difference between the total download volume and the total upload volume is used as the operating range of the download volume or the operating range of the upload volume of the merged station;

[0021] S32: If there are two or more sub - stations between any two compressor stations in the local pipeline section, according to the operating range of the download volume, the remaining sub - stations between these two compressor stations are merged with the sub - station with the largest operating range of the download volume in the local pipeline section topology, and the operating range of the download volume is accumulated;

[0022] S33: If there are two or more upload stations between any two compressor stations in the local pipeline section, according to the operating range of the upload volume, the remaining upload stations between these two compressor stations are merged with the upload station with the largest operating range of the upload volume in the local pipeline section topology, and the operating range of the upload volume is accumulated;

[0023] S34: If there are three or more sub - stations along the local pipeline section after the merging process in S31 - S33, two sub - stations with the largest operating range of the download volume are retained in the local pipeline section topology, the remaining sub - stations are merged with the nearest retained sub - station, and the operating range of the download volume is accumulated;

[0024] S35: If there are three or more upload stations along the local pipeline section after the merging process in S31 - S33, two upload stations with the largest operating range of the upload volume are retained in the local pipeline section topology, the remaining upload stations are merged with the nearest retained upload station, and the operating range of the upload volume is accumulated.

[0025] Furthermore, in step 4, the target response value of the response surface experiment scheme is the pipeline operation cost. The experimental factors include the inlet pressure, inlet flow rate of the compressor station at the starting point of the local pipeline section that can be merged, and the outlet pressure of the compressor station at the ending point. And according to the specific situation of the local pipeline section that can be merged, the experimental factors may also include the download volume of 0 to 2 sub - stations and the upload volume of 0 to 2 upload stations, and the level range of the experimental factors is determined by the operating condition range.

[0026] Furthermore, in step 6, the merging and simplification process is specifically to replace the local pipeline section with multiple compressor stations that can be merged with one merged compressor station in the global pipeline network topology, and further process the optimization model, including the following steps:

[0027] S61: In the objective function of the optimization model, replace the operation cost calculation formula of multiple compressor stations involved in the local pipeline section with the response model between the operating condition parameters and the operation cost established in step 5;

[0028] S62: In the constraint conditions of the optimization model, replace all the constraint conditions involved in the local pipeline section with the operation constraint conditions of one compressor station, including: compressor station flow constraint, compressor station inlet and outlet pressure constraints, and node flow balance constraint;

[0029] S63: In the decision variables of the optimization model, merge and replace the outlet pressure decision variables of multiple compressor stations involved in the local pipeline section with the outlet pressure decision variable of one compressor station.

[0030] Furthermore, in step 7, the approximate operating condition selection method includes the following steps:

[0031] S71: Adopt the operating condition parameters X of each scheme label in the multi-operating condition optimal operation scheme set of the local pipeline section label and the current operating condition parameters X current , calculate the Euclidean distance Dist considering data magnitude normalization as the distance evaluation index, as shown in formula (1):

[0032]

[0033] In the formula: Dist is the distance evaluation index, N local is the number of operating condition parameters of the local pipeline section, is the i-th parameter in the operating condition parameters of the scheme label, is the i-th parameter of the current operating condition parameters;

[0034] S72: According to the distance evaluation indexes of each scheme in the multi-operating condition optimal operation scheme set calculated in S71, select the scheme with the smallest distance evaluation index as the nearest scheme of the current operating condition.

[0035] Due to the above technical solutions adopted by the present invention, it has the following advantages: 1. By identifying the locally mergeable pipe sections of multiple compressor stations, the present invention constructs a response model between the operating condition parameters of the locally mergeable pipe sections and the operating cost, and then respectively performs merging and simplification processing on the global pipe network topology structure and the optimization model, so as to effectively reduce the dimension and simplify the problem of optimizing the operation of large-scale natural gas pipe networks, significantly reducing the search space and computational burden of the problem of optimizing the operation of large-scale pipe networks, greatly improving the solution speed, effectively supporting the rapid acquisition of the best operation plan for real-time operating conditions, and meeting the timeliness requirements of pipe network operation scheduling; 2. By summarizing the historical operation data of the pipe network, clarifying the operating condition range of the locally mergeable pipe sections of each multiple compressor station, and combining the response surface experimental design and analysis methods, the present invention establishes a set of best operation plans for multiple operating conditions of the locally mergeable pipe sections that can effectively cover the operating condition range, and proposes an approximate operating condition selection method, so as to quickly and reliably obtain the best operation plan of the locally mergeable pipe sections under different operating conditions; 3. By establishing a method for identifying and simplifying compressor stations that can be merged, the present invention can effectively identify and reasonably split the complex topology structure of the pipe network, and further delimit the locally mergeable pipe sections of multiple compressor stations according to the location distribution of the compressor stations, providing a reliable basis for subsequent merging and simplification processing, and ensuring the feasibility of finally solving the best operation plan; 4. By establishing a preliminary method for simplifying and merging processing, the present invention can reasonably simplify and merge the branch pipelines, distribution stations and uploading stations of the locally mergeable pipe sections according to the location distribution of the stations and the operating condition range, effectively reducing the number of experimental factors in the response surface design experiment and improving the operability of the response surface experimental plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0037] Figure 1 It is a schematic flow chart of a method for optimizing the operation of merging and simplifying pipe network compressor stations based on the response surface method provided by an embodiment of the present invention;

[0038] Figure 2 It is a schematic diagram of the global pipe network topology structure of a certain pipe network provided by an embodiment of the present invention

[0039] Figure 3 It is a schematic diagram of the identification result of the locally mergeable pipe sections of multiple compressor stations of a certain pipe network provided by an embodiment of the present invention

[0040] Figure 4Schematic diagram of the result after the combined local pipe segments of multiple compressor stations in a certain pipe network provided by the embodiment of the present invention are combined and simplified. Detailed implementation manners

[0041] The natural gas pipeline network consists of various types of stations and pipelines such as uploading stations, gas distribution stations, and compressor stations, and has a complex topological connection structure at the same time. This characteristic leads to the problem of optimizing the operation of large-scale natural gas pipeline networks not only having a high solution dimension, but also showing non-convex and non-linear characteristics, resulting in difficulties in quickly and reliably obtaining feasible and relatively optimal operation plans, being unable to adapt to the dynamic changes of the pipeline network operation conditions, and being difficult to meet the timeliness requirements of the pipeline network operation scheduling. For this reason, the present invention proposes a method for optimizing the operation of combined and simplified compressor stations in a pipeline network based on the response surface method. By combining the response surface experimental design and analysis technology, the problem of optimizing the operation of large-scale pipeline networks is reduced in dimension and simplified, significantly reducing the optimization space and calculation burden of the problem of optimizing the operation of large-scale pipeline networks, greatly improving the solution speed, effectively supporting the rapid acquisition of the best operation plan for real-time operation conditions, and meeting the timeliness requirements of the pipeline network operation scheduling.

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0043] The present invention provides a method for optimizing the operation of combined and simplified compressor stations in a pipeline network based on the response surface method, as Figure 1 shown, including the following steps:

[0044] Step 1: Collect the basic data of the pipeline network structure, construct the global pipeline network topology structure, and use the identification method for compressor stations that can be simplified and combined to identify the global pipeline network topology structure to obtain the combined local pipe segments of multiple compressor stations;

[0045] Specifically, the basic data of the pipeline network structure includes: pipeline number, pipeline flow direction, pipeline designed transportation capacity, pipeline start node and end node, node number, and node type.

[0046] In specific implementation, considering that the operating conditions of some compressor stations in local pipe sections of the natural gas pipeline network are only related to the operating conditions of the upstream and downstream compressor stations in the local pipe sections, the optimal operating schemes of some compressor stations in the local pipe sections can be solved in multiple operating conditions in advance, so as to reduce the optimization space dimension of the global pipeline network optimization problem. Therefore, through identification and splitting, the simplified merger compressor station identification method can select local pipe sections that can be merged and simplified from the complex topological structure of the large-scale pipeline network, laying a foundation for the dimensionality reduction and simplified processing of the global pipeline network operation optimization problem, including the following steps:

[0047] S11: According to the global pipeline network topological structure, pipeline flow direction and compressor station distribution, identify pipeline intersection nodes. If there is more than one pipeline along the pipelines connected by the intersection nodes with compressor stations, split the global pipeline network topological structure at the pipeline intersection nodes, traverse all pipeline intersection nodes, and obtain the local pipeline topological structure;

[0048] S12: Traverse and analyze the local pipeline topological structure obtained in S11. If the local pipeline topological structure contains two or more compressor stations, further delimit and obtain the multi-compressor-station mergeable local pipe section with the first compressor station node upstream of the local pipeline as the starting point and the last compressor station downstream of the local pipeline as the end point.

[0049] Step 2: Collect the basic operating data of the pipeline network and determine the operating condition range of each multi-compressor-station mergeable local pipe section;

[0050] Specifically, the basic pipeline operation data includes: pipeline length, pipeline diameter, pipeline roughness, maximum designed operating pressure of the pipeline, allowable pressure range for downloading at the distribution station, allowable flow range for passing through the compressor station, allowable pressure range for inlet and outlet of the compressor station, and historical operating data of pressure and flow of each node and pipeline in the pipeline network.

[0051] Specifically, the operating condition range of the multi-compressor-station mergeable local pipe section includes: the inlet pressure range of the starting compressor station, the inlet flow range of the starting compressor station, the outlet pressure range of the ending compressor station, as well as the downloading amount range of the distribution stations along the pipe section and the uploading amount range of the uploading stations.

[0052] Step 3: Traverse and analyze each multi-compressor-station mergeable local pipe section. If there are branch pipelines or more than two distribution stations or uploading stations along the local pipe section, according to the station location distribution and operating condition range, adopt the preliminary simplified merger processing method to simplify the branch pipelines and approximately merge multiple distribution stations or uploading stations to obtain the multi-compressor-station mergeable local pipe section after preliminary simplified merger;

[0053] In specific implementation, since the compressor stations are mainly the optimization objects in the operation optimization of natural gas pipeline networks, the preliminary simplified merging method can reasonably simplify and merge the branch pipelines, gas distribution stations, and uploading stations of local pipe segments according to the location distribution and operating condition ranges, effectively reducing the number of experimental factors in the response surface design experiment, including the following steps:

[0054] S31: If there are branch pipelines along the local pipe segment, the operating ranges of the downloaded volumes of the gas distribution stations along the branch pipelines and the operating ranges of the uploaded volumes of the uploading stations are respectively accumulated. If the total downloaded volume is larger, all the stations along the branch pipeline are merged into one gas distribution station in the topological structure of the local pipe segment; otherwise, if the total uploaded volume is larger, they are merged into one uploading station. The merged station is set at the connection of the branch pipeline and the local pipe segment, and the absolute value of the difference between the total downloaded volume and the total uploaded volume is used as the operating range of the downloaded volume or the operating range of the uploaded volume of the merged station;

[0055] S32: If there are two or more gas distribution stations between any two compressor stations in the local pipe segment, according to the operating range of the downloaded volume, the remaining gas distribution stations between these two compressor stations are merged with the gas distribution station with the largest operating range of the downloaded volume in the topological structure of the local pipe segment, and the operating ranges of the downloaded volumes are accumulated;

[0056] S33: If there are two or more uploading stations between any two compressor stations in the local pipe segment, according to the operating range of the uploaded volume, the remaining uploading stations between these two compressor stations are merged with the uploading station with the largest operating range of the uploaded volume in the topological structure of the local pipe segment, and the operating ranges of the uploaded volumes are accumulated;

[0057] S34: If there are still three or more gas distribution stations along the local pipe segment after the merging process in S31 - S33, two gas distribution stations with the largest operating ranges of the downloaded volume are retained in the topological structure of the local pipe segment, and the remaining gas distribution stations are merged with the nearest retained gas distribution station, and the operating ranges of the downloaded volumes are accumulated;

[0058] S35: If there are still three or more uploading stations along the local pipe segment after the merging process in S31 - S33, two uploading stations with the largest operating ranges of the uploaded volume are retained in the topological structure of the local pipe segment, and the remaining uploading stations are merged with the nearest retained uploading station, and the operating ranges of the uploaded volumes are accumulated.

[0059] Step 4: For each locally mergeable pipe segment with multiple compressor stations, combined with the operating condition parameters and ranges, establish the corresponding factor level table, and use the Box - Behnken method to design a response surface experiment scheme including multiple groups of different combinations of operating condition parameters;

[0060] In specific implementation, the target response value of the response surface experiment plan is the operating cost of the local pipeline section. The experimental factors include the inlet pressure of the starting compressor station of the local pipeline section that can be combined by multiple compressor stations, the outlet pressure of the ending compressor station, and the outlet flow rate. According to the specific situation of the local pipeline section that can be combined by multiple compressor stations, the experimental factors may also include the download volume of 0 to 2 delivery stations and the upload volume of 0 to 2 loading stations. In addition, the level range of the experimental factors is determined by the operating condition range.

[0061] Specifically, the parameters in the operating condition combination include the experimental factors involved in the combinable local pipeline section, specifically, the value combination of the experimental factors within the operating condition range of the combinable local pipeline section.

[0062] Step 5: Use the optimization model for minimizing the pipeline network operating cost. For each local pipeline section that can be combined by multiple compressor stations, solve the optimal operating plan and the minimum operating cost of each operating condition parameter combination in the response surface experiment plan, construct a set of optimal operating plans for multiple conditions of the local pipeline section, and fit the response model between the operating condition parameters and the operating cost of the local pipeline section.

[0063] In specific implementation, the optimization model for minimizing the pipeline network operating cost aims to minimize the sum of the operating costs of the compressor stations along the pipeline. Considering the pressure and flow constraints of the stations along the pipeline and the operating constraints of the pipeline itself, it solves the combination plan of the outlet pressures of the compressor stations with the lowest operating cost. The optimization model for minimizing the pipeline network operating cost includes the objective function for minimizing the pipeline network operating cost, constraint conditions, and decision variables:

[0064] S51: The objective function for minimizing the pipeline network operating cost is to minimize the sum of the operating costs of the compressor stations along the pipeline, as shown in Equation (2):

[0065]

[0066] In the formula: F is the total operating cost of the pipeline, i is the i-th compressor station along the pipeline, N c is the number of compressor stations along the pipeline, and are the operating characteristic parameters of the i-th compressor station, is the inlet pressure of the i-th compressor station, is the inlet flow rate of the i-th compressor station, is the outlet pressure of the i-th compressor station, η i is the operating efficiency of the i-th compressor station, pr i energy is the unit energy consumption economic cost of the i-th compressor station;

[0067] S52: The constraint conditions include: pipeline pressure constraint, pipeline gas flow rate constraint, pipeline hydraulic pressure drop constraint, upload pressure constraint at the upload station, download pressure constraint at the distribution station, flow rate constraint at the compressor station, inlet and outlet pressure constraints at the compressor station, and node flow balance constraint;

[0068] Specifically, the pipeline pressure constraint characterizes the starting point of the pipeline and the ending point pressure being less than the maximum design pressure of the pipeline As shown in Equation (3):

[0069]

[0070] Specifically, the pipeline gas flow rate constraint characterizes the gas flow rate inside the pipeline being less than the maximum erosion flow rate As shown in Equation (4):

[0071]

[0072] Specifically, the pipeline hydraulic pressure drop constraint characterizes that the pressures at both ends of the pipeline nodes should satisfy the hydraulic pressure drop equation relationship, as shown in Equation (5):

[0073]

[0074] In the formula: is the mass flow rate of pipeline (i, j), kg / s, and are the starting point and ending point pressures of pipeline (i, j) respectively, Pa, D (i,j) is the inner diameter of pipeline (i, j), m, λ (i,j) is the hydraulic friction coefficient of pipeline (i, j), Z is the natural gas compression factor; R is the natural gas gas constant, T (i,j) is the average temperature of natural gas inside pipeline (i, j), K, L (i,j) is the length of pipeline (i, j), m.

[0075] The present invention uses the Colebrook-White formula to calculate the friction coefficient, and this method has the advantage of high precision, as shown in Equation (6):

[0076]

[0077] In the formula: ε (i,j) is the absolute roughness of pipeline (i, j), m, and Re is the Reynolds number of the fluid inside pipeline (i, j).

[0078] Specifically, the upload pressure constraint at the upload station characterizes that the upload pressure at the upload station meets the maximum upload pressure and the minimum upload pressure is restricted as shown in Equation (7):

[0079]

[0080] Specifically, the download pressure constraint of the distribution station characterizes the download pressure of the distribution station satisfies the maximum upload pressure and the minimum upload pressure is restricted as shown in Equation (8):

[0081]

[0082] Specifically, the flow rate constraint of the compressor station characterizes the flow rate passing through the compressor station satisfies the processing capacity limit of the in-station equipment as shown in Equation (9):

[0083]

[0084] Specifically, the inlet and outlet pressure constraints of the compressor station characterize the inlet outlet pressure satisfying the operating pressure limit of the compressor as shown in Equation (10):

[0085]

[0086] Specifically, the node flow balance constraint characterizes that according to the law of conservation of mass, the inflow at any node should be equal to the outflow, as shown in Equation (11):

[0087]

[0088] In the formula: is the absolute volume flow rate between the i-th node and the j-th component, m 3 / s, α (i,j) is used to characterize the flow direction between the i-th node and the j-th component. If it flows from the j-th component to the i-th node, it is -1. If it flows from the i-th node to the j-th component, it is 1. U i is the set of components connected to the i-th node.

[0089] S53: The decision variables include: the outlet pressure of the i-th compressor station along the pipeline as shown in Equation (12):

[0090]

[0091] In specific implementation, by adopting the above optimization model for minimizing the operation cost of the pipe network and combining heuristic optimization algorithms such as genetic algorithms or particle swarm algorithms, multiple optimization models can be correspondingly established for each combinable local pipe section of multiple compressor stations, and the optimal operation scheme can be solved under different working conditions.

[0092] Specifically, the set of optimal operation schemes for multiple working conditions of the local pipe section is stored with the operation condition parameters as labels, and the optimal operation scheme obtained correspondingly by using the optimization model for minimizing the operation cost of the pipe network under these condition parameters.

[0093] Step 6: Adopt the optimization model for minimizing the operation cost of the pipeline, solve the global pipe network operation optimization problem, and combine the response model established in Step 5 to perform merging and simplification processing on each combinable local pipe section of multiple compressor stations in the global pipe network topology and optimization model, and use the optimization algorithm to solve and obtain the optimal operation scheme of the global pipe network;

[0094] In specific implementation, the merging and simplification processing is specifically to replace the combinable local pipe section of multiple compressor stations in the global pipe network topology with a merged compressor station, and further simplify the optimization model, so as to improve the solution efficiency, including the following steps:

[0095] S61: In the objective function of the optimization model, replace the operation cost calculation formulas of multiple compressor stations involved in the local pipe section with the response model between the condition parameters and the operation cost established in Step 5;

[0096] S62: In the constraint conditions of the optimization model, replace all the constraint conditions involved in the local pipe section with the operation constraint conditions of one compressor station, including: compressor station flow constraint, compressor station inlet and outlet pressure constraints, and node flow balance constraint;

[0097] S63: In the decision variables of the optimization model, merge and replace the outlet pressure decision variables of multiple compressor stations involved in the local pipe section with the outlet pressure decision variable of one compressor station.

[0098] Step 7: According to the optimal operation scheme of the global pipe network in Step 6, determine the current operation condition parameters of each combinable local pipe section of multiple compressor stations, and adopt the approximate condition selection method to correspondingly select the most similar scheme from the set of optimal operation schemes for multiple working conditions of the local pipe section as the optimal operation scheme of the current local pipe section;

[0099] In specific implementation, the approximate condition selection method can compare the operation condition parameters of the current local pipe section with the labels of each scheme in the set of optimal operation schemes for multiple working conditions of the local pipe section, and select the scheme with the most similar conditions as the optimal operation scheme of the current local pipe section, including the following steps:

[0100] S71: Adopt the optimal operation scheme for multiple working conditions of local pipe segments to centralize the operation condition parameters X of each scheme label label and the current operation condition parameters X current , calculate the Euclidean distance Dist considering data magnitude normalization as the distance evaluation index, as shown in Equation (1).

[0101] S72: According to the distance evaluation indexes of each scheme in the optimal operation scheme set for multiple working conditions of local pipe segments calculated in S71, select the scheme with the smallest distance evaluation index as the nearest scheme for the current working condition.

[0102] Step 8: After completing Steps 1 - 5, when the operation condition of the pipe network changes, according to the latest operation condition parameters, re - perform Steps 6 and 7, and the optimal operation scheme with the lowest operation cost under the real - time working condition can be quickly solved and obtained.

[0103] In specific implementation, Steps 1 - 5 have completed the establishment of the optimal operation scheme set for multiple working conditions of local pipe segments of the local pipe segments that can be merged by multiple compressor stations, as well as the merging and simplification processing of the global pipe network topology structure and the optimization model. When the operation condition changes, the optimal operation scheme for the real - time working condition can be directly and quickly sought, which can effectively meet the timeliness requirements of the pipe network operation scheduling.

[0104] Example:

[0105] Adopt an operation optimization method for merging and simplifying compressor stations in a pipe network based on the response surface method proposed by the present invention. Take a long - distance natural gas pipe network as an example, combine the relevant data required for the implementation of the method, and conduct multi - aspect evaluations around the solution time and optimization effect to further illustrate the present invention, and at the same time verify the reliability and effectiveness of the present invention.

[0106] A long - distance natural gas pipe network has 4048 km of pipelines, including 6 loading stations, 27 off - take stations and 12 compressor stations along the line, and has a complex pipe network topology structure. The global pipe network topology structure of this pipe network is as Figure 2 shown. Use the compressor station mergeable identification method to identify the global pipe network topology structure, and obtain the local pipe segments that can be merged by multiple compressor stations as Figure 3 shown. Among them, 4 pipeline intersection nodes are identified in this pipe network and marked with solid circles in Figure 3 , and 10 local pipeline topology structures are obtained after splitting and marked with solid rectangles in Figure 3 , and further delimit 3 local pipe segments that can be merged by multiple compressor stations and marked with dashed rectangles in Figure 3 .

[0107] By collecting the basic operation data of the pipeline network, determine the operating condition range of the local pipeline sections that can be merged by each multi-compressor station. The starting compressor station of the local pipeline section 1 that can be merged by the multi-compressor station is Compressor Station 1, and the ending compressor station is Compressor Station 4. There is 1 uploading station and 8 gas distribution stations along the line, and the operating condition range is shown in Table 1.

[0108] Table 1 Operating Condition Range of the Local Pipeline Section 1 that can be Merged by the Multi-compressor Station

[0109] Project Operating condition range Project Operating condition range Inlet pressure of compressor station 1 (MPa) (6,8) <![CDATA[Download volume of the distribution station 5 (m 3 / s)]]> (82,123) Outlet pressure of compressor station 4 (MPa) (9,10.5) <![CDATA[Download volume of the distribution station 6 (m 3 / s)]]> (85,127) <![CDATA[Outlet flow rate of compressor station 4 (m 3 / s)]]> (152,228) <![CDATA[Download volume of the distribution station 7 (m 3 / s)]]> (48,72) <![CDATA[Upload amount of Upload Station 2 (m 3 / s)]]> (350,530) <![CDATA[Download volume of the distribution station 8 (m 3 / s)]]> (121,182) <![CDATA[Download volume of Sub - station 3 (m 3 / s)]]> (74,112) <![CDATA[Download volume of the branch transfer station 9 (m 3 / s)]]> (129,193) <![CDATA[Download volume of the distribution station 4 (m 3 / s)]]> (63,95)

[0110] The starting compressor station of the local pipeline section 2 that can be merged by the multi-compressor station is Compressor Station 6, and the ending compressor station is Compressor Station 7. There are 3 gas distribution stations along the line, and the operating condition range is shown in Table 2.

[0111] Table 2 Operating Condition Range of the Local Pipeline Section 2 that can be Merged by the Multi-compressor Station

[0112] Project Operating condition range Project Operating condition range Inlet pressure of compressor station 6 (MPa) (4.5,6) <![CDATA[Download volume of the distribution station 12 (m 3 / s)]]> (96,144) Outlet pressure of compressor station 7 (MPa) (6,7) <![CDATA[Download volume of the distribution station 13 (m 3 / s)]]> (44,66) <![CDATA[Outlet flow rate of compressor station 7 (m 3 / s)]]> (102,154) <![CDATA[Download volume of the distribution station 14 (m 3 / s)]]> (20,30)

[0113] The starting compressor station of the local pipeline section 3 that can be merged by the multi-compressor station is Compressor Station 8, and the ending compressor station is Compressor Station 10. There are 2 uploading stations and 3 gas distribution stations along the line, and the operating condition range is shown in Table 3.

[0114] Table 3 Operating Condition Range of the Local Pipeline Section 3 that can be Merged by the Multi-compressor Station

[0115] Project Operating condition range Project Operating condition range Inlet pressure of compressor station 8 (MPa) (4.5,6) <![CDATA[Upload amount of upload station 5 (m 3 / s)]]> (110,156) Outlet pressure of compressor station 10 (MPa) (6,7) <![CDATA[Download volume of the distribution station 17 (m 3 / s)]]> (51,77) <![CDATA[Outlet flow rate of compressor station 10 (m 3 / s)]]> (119,178) <![CDATA[Download volume of the distribution station 18 (m 3 / s)]]> (63,95) <![CDATA[Upload amount of upload station 4 (m 3 / s)]]> (88,132) <![CDATA[Download volume of the distribution station 19 (m 3 / s)]]> (46,69)

[0116] For the case where there are branch pipelines or multiple uploading stations and gas distribution stations along the local pipeline sections that can be merged by the multi-compressor station, according to the location distribution of the stations and the operating condition range, adopt the preliminary approximate merging method to approximately merge the branch pipelines and multiple gas distribution stations or uploading stations to obtain the preliminarily merged local pipeline sections that can be merged by the multi-compressor station. In the local pipeline section 1 that can be merged by the multi-compressor station, after the approximate merging process, there is 1 uploading station of the local pipeline section and 2 gas distribution stations of the local pipeline section. In the local pipeline section 2 that can be merged by the multi-compressor station, after the approximate merging process, there is 1 gas distribution station of the local pipeline section. In the local pipeline section 3 that can be merged by the multi-compressor station, after the approximate merging process, there are 2 uploading stations of the local pipeline section and 2 gas distribution stations of the local pipeline section.

[0117] Based on the preliminarily merged local pipeline sections that can be merged by the multi-compressor station, further establish the factor level tables of each local pipeline section that can be merged by the multi-compressor station, as shown in Tables 4, 5, and 6.

[0118] Table 4 Factor Level Table of the Local Pipeline Section 1 that can be Merged by the Multi-compressor Station

[0119]

[0120] Table 5 Factor level table of local pipe section 2 that can be combined by multiple compressor stations

[0121]

[0122]

[0123] Table 6 Factor level table of local pipe section 3 that can be combined by multiple compressor stations

[0124]

[0125] Based on the factor level tables of local pipe sections that can be combined by each multiple compressor station, the Box-Behnken method is used to combine within the operating range of operating parameters, forming multiple combinations of operating parameter values. Accordingly, 3 response surface experimental schemes are established. Using the optimization model of minimizing the pipeline network operating cost, corresponding optimization models are established for each local pipe section that can be combined by multiple compressor stations, and the best operating scheme is solved for each operating condition in the response surface experimental scheme, accordingly forming 3 sets of best operating schemes for multiple operating conditions of local pipe sections. Among them, the set of best operating schemes for multiple operating conditions of local pipe section 1 that can be combined by multiple compressor stations contains 54 best operating schemes for different operating conditions, the set of best operating schemes for multiple operating conditions of local pipe section 2 that can be combined by multiple compressor stations contains 29 best operating schemes for different operating conditions, and the set of best operating schemes for multiple operating conditions of local pipe section 3 that can be combined by multiple compressor stations contains 62 best operating schemes for different operating conditions.

[0126] Meanwhile, response models between the operating condition parameters (X) and operating cost (Y) of the local pipe section are constructed based on the response surface experimental scheme, as shown in Equations (13), (14) and (15) respectively.

[0127]

[0128]

[0129] The response models have high calculation accuracy. Through variance analysis, the P values of the above model equations are all less than 0.0001, and the Rs 2 are 0.9982, 0.9999 and 0.9951 respectively, indicating that the response surface experimental scheme designed based on the operating condition range of local pipe sections that can be combined by multiple compressor stations can effectively cover the operating condition range, and the constructed response models can accurately calculate the operating cost of local pipe sections that can be combined by multiple compressor stations according to different operating condition parameters.

[0130] Further, an optimization model for minimizing the operation cost of the pipeline network is adopted to solve the global pipeline network operation optimization problem. Combining the established response model, in the optimization model, the locally mergeable pipe segments of each multi-compressor station are merged and simplified. In the global pipeline network topology, the locally mergeable pipe segments of the multi-compressor station are replaced with a merged compressor station as Figure 4 shown, and corresponding processing is carried out in the objective function, constraint conditions, and decision variables of the global pipeline network optimization model. After the merging and simplification process, the number of decision variables of the global pipeline network optimization model is reduced from 12 to 6, and the number of decision variables decreases by 50%. The number of constraint conditions of the optimization model is reduced from 237 to 113, and the number of constraint conditions decreases by 52.32%. It can be seen that the merging and simplification method proposed by the present invention can effectively reduce the scale and optimization dimension of the optimization model and reduce the solution complexity of the optimization model.

[0131] The global pipeline network optimization model after the merging and simplification process can quickly solve to obtain the best operation plan for the simplified global pipeline network. Further, by determining the current operation condition parameters of each locally mergeable pipe segment of the multi-compressor station and adopting the approximate operation condition selection method, the most approximate plan is correspondingly selected from the set of best operation plans for multiple operating conditions of the local pipe segment as the best operation plan for the current local pipe segment, so as to quickly establish the best operation plan for the global pipeline network.

[0132] Further, based on the above-mentioned long-distance natural gas pipeline network, taking the actual operation conditions as a case, the global pipeline network optimization model without simplification and the pipeline compressor station merging and simplification operation optimization method proposed by the present invention are respectively used, combined with various optimization algorithms, to solve the best operation plan for the global pipeline network, and the method proposed by the present invention is verified and analyzed from two aspects of solution time and optimization effect. The specific parameters of the actual operation condition case are shown in Table 7.

[0133] Table 7 Specific parameters of the actual operation condition case

[0134]

[0135]

[0136] Based on the above-mentioned operation condition case, for the global pipeline network optimization model without simplification and the pipeline compressor station merging and simplification operation optimization method proposed by the present invention, the particle swarm optimization algorithm and the genetic algorithm are respectively used for solution. Under the same computer hardware conditions, the solution optimization results are shown in Table 8.

[0137] Table 8 Comparison of solution optimization results

[0138]

[0139] Compared with the global pipeline network optimization model without simplification, the proposed method for optimizing the combined and simplified operation of pipeline network compressor stations in the present invention reduces the required solution time by 34.27% and 64.97% respectively when using the genetic algorithm and the particle swarm algorithm, and the operating costs only increase by 0.95% and 0.23% respectively. This verifies that the proposed method for optimizing the combined and simplified operation of pipeline network compressor stations based on the response surface method in the present invention can effectively reduce the required solution time for the optimal operation plan of the natural gas pipeline network, and at the same time has a good optimization effect. In addition, the method proposed in the present invention has good promotion potential, has strong reliability and practicability for the rapid solution of the operation optimization of large-scale natural gas pipeline networks. It can effectively reduce the solution time of the optimal operation plan of the pipeline network through the pre-establishment of the optimal operation plan set of multiple working conditions for local pipe segments and the pre-simplification of the optimization model, and can effectively cover the range of operating conditions of local pipe segments in combination with the response surface experimental design method to ensure the feasibility of the solution operation plan, providing fast and reliable decision support for the real-time operation scheduling of the pipeline network.

Claims

1. A simplified operation optimization method for pipeline compressor station merger based on response surface methodology, characterized in that The following steps are involved: Step 1: Collect basic data of the pipe network structure, construct the global pipe network topology, use the simplified mergeable compressor station identification method to identify the global pipe network topology, and obtain the mergeable local pipe sections of multiple compressor stations; Step 2: Collect basic operation data of the pipeline network and determine the operating range of the local pipeline sections that can be merged for each multi-compressor station; Step 3: Traverse and analyze the local pipeline sections that can be merged for each multi-compressor station. If there are branch pipelines or more than two sub-transmission stations or upload stations along the local pipeline section, then according to the station location distribution and operating condition range, adopt a preliminary simplified merging method to simplify the branch pipelines and approximately merge multiple sub-transmission stations or upload stations to obtain the local pipeline sections that can be merged for the multi-compressor stations after preliminary simplification and merging; Step 4: For each multi-compressor station that can merge local pipe sections, the corresponding factor level table is established in combination with the operating condition parameters and ranges, and the Box-Behnken method is used to design a response surface experimental plan containing multiple groups of different operating condition parameter combinations; Step 5: Adopt the pipeline network operation cost minimization optimization model, for each multi-compressor station that can merge local pipe sections, solve the optimal operation scheme and minimum operation cost of each combination of operating condition parameters in the response surface experiment scheme, build the optimal operation scheme set of multiple operating conditions of the local pipe section, and fit the response model between the local pipe section operating parameters and the operation cost; Step 6: Use the pipeline network operation cost minimization optimization model to solve the global pipeline network operation optimization problem, and use the response model established in step 5 to merge and simplify the local pipe sections that can be merged in each multi-compressor station in the global pipeline network topology structure and optimization model, and combine the optimization algorithm to solve and obtain the best operation plan for the global pipeline network; Step 7: According to the optimal operation plan of the global pipeline network in step 6, the current operating condition parameters of the local pipeline section that can be merged in each multi-compressor station are determined, and the approximate operating condition selection method is adopted to select the most approximate plan from the set of optimal operation plans of multiple operating conditions of the local pipeline section as the optimal operation plan of the current local pipeline section; Step 8: After completing steps 1 to 5, when the operating conditions of the pipeline network change, repeat steps 6 and 7 according to the latest operating condition parameters to quickly obtain the optimal operating plan with the lowest operating cost under real-time conditions.

2. A method for optimizing the operation of pipeline network compressor stations based on response surface methodology according to claim 1, characterized in that: In step 1 of claim 1, the compressor station identification method may include the following steps: S11: According to the global pipeline network topology, pipeline flow direction and compressor station distribution, identify the pipeline intersection node. If there is a compressor station along one or more pipelines connected to the intersection node, split the global pipeline network topology at the pipeline intersection node, traverse all pipeline intersection nodes, and obtain the local pipeline topology structure; S12: traverse and analyze the local pipeline topology structure obtained in S11. If the local pipeline topology structure contains two or more compressor stations, the node of the first compressor station upstream of the local pipeline is taken as the starting point and the node of the last compressor station downstream of the local pipeline is taken as the end point, and the local pipe section that can be merged with multiple compressor stations is further delineated; In step 3 of claim 1, the preliminary simplified merging processing method comprises the following steps: S31: If there is a branch pipeline along the local pipeline section, the download volume operation range of the sub-transmission station and the upload volume operation range of the upload station along the branch pipeline are accumulated respectively. If the total download volume is larger, all stations along the branch pipeline are merged into one sub-transmission station in the local pipeline section topology structure. Conversely, if the total upload volume is larger, they are merged into one upload station. The merged station is set at the connection between the branch pipeline and the local pipeline section. The absolute value of the difference between the total download volume and the total upload volume is used as the download volume operation range or the upload volume operation range of the merged station. S32: If there are two or more sub-transmission stations between any two compressor stations in the local pipeline section, then according to the download volume operation range, the remaining sub-transmission stations between the two compressor stations are merged with the sub-transmission station with the largest download volume operation range in the local pipeline section topology structure, and the download volume operation range is accumulated; S33: if there are two or more upload stations between any two compressor stations in the local pipe section, then according to the upload capacity operation range, the remaining upload stations between the two compressor stations are merged with the upload station with the largest upload capacity operation range in the local pipe section topology structure, and the upload capacity operation range is accumulated; S34: If there are still three or more sub-transmission stations along the local pipeline section after the merging process of S31-S33, the two sub-transmission stations with the largest download volume operation range are retained in the local pipeline section topology structure, and the remaining sub-transmission stations are merged with the nearest reserved sub-transmission stations, and the download volume operation range is accumulated; S35: If there are still three or more upload stations along the local pipe section after the merging process of S31-S33, the two upload stations with the largest upload volume operation range are retained in the local pipe section topology structure, the remaining upload stations are merged with the nearest reserved upload station, and the upload volume operation range is accumulated; In step 4 of claim 1, the target response value of the response surface experiment scheme is the pipeline operation cost, and the experimental factors include the inlet pressure and inlet flow of the starting compressor station of the mergeable local pipeline section and the outlet pressure of the terminal compressor station, and according to the specific conditions of the mergeable local pipeline section, the experimental factors may also include the download volume of 0 to 2 sub-transmission stations and the upload volume of 0 to 2 upload stations, and the level range of the experimental factors is determined by the operating condition range; In step 6 of claim 1, the merging and simplifying process is specifically to replace the mergeable local pipe sections of multiple compressor stations with one merged compressor station in the global pipe network topology structure, and further process the optimization model, including the following steps: S61: In the objective function of the optimization model, the operating cost calculation formula of the multiple compressor stations involved in the local pipe section is replaced by the response model between the operating parameters and the operating cost established in step 5 of claim 1; S62: In the constraints of the optimization model, all constraints involved in the local pipe section are replaced with operation constraints of a compressor station, including: compressor station flow constraint, compressor station inlet and outlet pressure constraints, and node flow balance constraint; S63: Among the decision variables of the optimization model, the outlet pressure decision variables of multiple compressor stations involved in the local pipeline section are merged and replaced with the outlet pressure decision variable of one compressor station.

Images