A method for optimizing the layout of pipe diameters in urban ring water supply networks
By establishing a pipe database and using an adaptive spiral Lévy flight iterative optimization algorithm, the pipe diameter design of the ring water supply network was optimized, solving the problems of investment waste and water supply reliability in traditional designs, and achieving the construction of the network with the lowest investment.
Patent Information
- Application Number
- CN202110266476.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-03-12
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2041-03-12
AI Technical Summary
In existing technologies, the pipe diameter design of ring-shaped water supply networks mainly relies on experience, resulting in wasted investment and difficulty in ensuring water supply reliability. Existing heuristic algorithms cannot effectively optimize this.
By establishing a local pipe database, utilizing the EPANET and ICS-EPANET algorithms, and combining adaptive spiral Lévy flight and iterative optimization methods based on dynamic discovery probability, a pipe diameter optimization model that interacts with the local pipe database is generated to optimize the pipe diameter combination of the pipeline network system to minimize investment costs.
It achieves a reduction in the total investment cost of pipeline construction while ensuring the safety and reliability of water supply, avoids the local optima and inaccuracies of traditional methods, and is suitable for pipeline design of complex systems.
Smart Images

Figure CN112966359B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of municipal engineering and urban water supply pipe network, and relates to a pipe diameter optimization layout method, in particular to a pipe diameter optimization layout method for a ring-shaped urban water supply pipe network. Background Art
[0002] Ring and tree-shaped water supply networks are important basic investments in the process of urban modernization. Compared with tree-shaped networks, ring-shaped networks have become the first choice for the design of urban water supply systems due to their strong terrain adaptability, high water supply reliability, and convenient maintenance. However, the construction of ring-shaped networks is very expensive. The traditional design method of determining the pipe diameter based on experience will cause a waste of investment. If the optimization method with the goal of reducing investment costs is used for pipe diameter design, it can not only save a lot of investment costs, but also make the water supply reliability of the water supply network higher, which can not only ensure the water supply requirements of each demand point, but also prevent the water supply pressure of the network from being wasted. With the continuous increase in urban water supply demand and the increasing complexity of urban water supply systems, how to develop and apply optimization algorithms to reduce investment costs as much as possible while ensuring water supply safety and reliability has become a research hotspot and key issue in current pipe network design. In recent years, science and technology have continued to advance, and computer technology and artificial intelligence algorithms have continued to develop, which has greatly promoted the development of heuristic algorithms, and the idea of using computer technology to deal with water supply network optimization design problems has begun to be proposed.
[0003] At present, the determination of water supply network diameters in actual projects mainly relies on the experience of designers, which cannot achieve truly accurate optimization. In addition, various heuristic algorithms have their own advantages and disadvantages, and cannot be directly used for water supply network optimization design problems. Summary of the invention
[0004] The purpose of the present invention is to address the deficiencies in the above-mentioned prior art and to propose a method for optimizing the layout of pipe diameters of urban ring water supply networks. Through this optimization method, the pipe diameters of the network can be optimized while ensuring the safety and reliability of water supply, so as to minimize the total construction investment cost.
[0005] The technical solution of the present invention is a method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network, characterized in that it comprises the following steps:
[0006] (1) Obtain commercial information on local water supply pipes of different pipe materials and establish a local pipe database;
[0007] (2) Based on the hydraulic analysis engine EPANET, the initial conditions and established layout of the urban ring water supply pipe network system are improved, and the .net file of the urban ring water supply pipe network system is output; the pipe network system .net file is read through the C language, and attribute categories are added to each data layer of the initial conditions and established layout of the pipe network system, and the .inp file is created and the attribute categories and their corresponding pipe network information are written;
[0008] (3) Read the .inp file of the pipe network system in binary mode, automatically extract the data information required for the pipe diameter optimization model in the .inp file according to the attribute category, and automatically generate a pipe diameter optimization model that interacts with the local pipe database in combination with the local pipe database;
[0009] (4) The pipe diameter matrix U of the water supply network system is improved through the EPANET deep coupling cuckoo algorithm program (ICS-EPANET), which completes the automatic iterative optimization of the pipe diameter matrix U of the pipe network system and the information feedback of the local pipe database, and determines the pipe diameter combination with the lowest investment value of the pipe network construction. The specific ICS-EPANET algorithm steps are as follows:
[0010] (4.1) Initialize ICS-EPANET algorithm parameters: set the number of rows (N) of the pipe diameter matrix U, the upper limit of the probability of discovery (Pa max ), lower limit of discovery probability (Pa min ), maximum step size control factor (α max ) and the maximum number of iterations (t max );
[0011] (4.2) t = 1, the initial N × n diameter matrix U = (d ij ) N×n ;
[0012] (4.3) In the tth iteration, ICS-EPANET is used to adaptively iterate U, and its iterative formula is as follows:
[0013] Updated method for adaptive spiral Lévy flight:
[0014]
[0015] Dynamic discovery probability update method:
[0016] x i t+1 =x i t +γ·H(Pa max -(Pa max -Pa min )t / t max -ε)·(xj t -x k t ) (2)
[0017] Where: x i t 、x i t+1 are the i-th matrix row in the t-th and t+1-th generation pipe diameter matrices U respectively; α is the step size control factor, calculated using formula (3); Lévy(λ) is the random search path matrix, which obeys the Lévy distribution: Lévy(λ)=μ·Φ / |ν| 1 / β , μ and ν follow the standard normal distribution, Φ = [Γ(1+β)·sin(π·β / 2)] / [Γ(1+β)·β·2 (β-1) / 2 / 2], Γ(1+β) is the Γ function, that is β is a constant, β∈[0,2]; x best t is the optimal pipe diameter combination of the tth generation, D t For x i t to x best t The distance, D t =|x best t -x i t |, b is the spiral constant, p and l are random numbers between [0,1], H is the Heaviside function, ε is a random number that follows the generalized Pareto distribution, and ε=(0.83+0.139·x) -6 , x, γ are random numbers that obey uniform distribution, γ∈[0,1], x j t 、x k t are the two random matrix rows in the t-th generation pipe diameter matrix U;
[0018]
[0019] (4.4) The optimized pipe diameter matrix U is discretely valued through the approximation principle of step (4.2), and each matrix row of the new pipe diameter matrix U covers the "pipe diameter" attribute category in the .inp file, and is fed back with the local pipe database information, and the Hezen-Williams coefficient in the .inp file is adjusted to perform EPANET hydraulic judgment and discard the matrix rows that do not meet the constraints; then the new matrix rows are supplemented through the random walk method and approximation principle of step (4.2), and the new matrix rows are again subjected to EPANET hydraulic judgment through the .inp file until all matrix rows of the pipe diameter matrix U meet the constraints, and the pipeline network construction investment values of all matrix rows in the pipe diameter matrix U are calculated and sorted, t = t + 1;
[0020] (4.5)t>t max Terminate the iteration and determine the optimal pipe diameter matrix row;
[0021] (5) After the ICS-EPANET optimization of the pipe diameter of the pipe network system is completed, all attribute categories and their corresponding pipe network information in the .inp file corresponding to the row of the optimal pipe diameter matrix are output.
[0022] The pipe database described in step (1) includes: the inner diameters of ductile iron pipes, PE pipes, welded steel pipes, and concrete pipes, and their corresponding commercial prices and Hezen-Williams coefficients.
[0023] The initial conditions and established layout of the urban ring water supply network system described in step (2) include: the established number of each node, the established elevation of each node, the established outflow of each node, the established number of each pipe section, the established length of each pipe section, the initial pipe diameter of each pipe section, the initial Hezen-William coefficient of each pipe section, the optional commercial pipe diameters and their corresponding prices and Hezen-William coefficients.
[0024] The attribute categories of the data information described in step (2) include: pipe segment number, pipe diameter, pipe length, Hazen-Williams coefficient, node number and node demand.
[0025] The optimization model described in step (3) is based on the diameter of each pipe segment (D i ) is a discrete decision variable, and the total cost of pipeline network investment is minimized (C min ) as the goal, and the node continuity equation, ring energy equation and node minimum water head are used as constraints; and the optimization model D i Form information feedback with the local pipe database, and i Automatically adjust the actual unit price and Hezen-Williams coefficient of the pipeline in the objective function and constraint conditions. The specific optimization model is as follows:
[0026] Objective function:
[0027] Constraint 1 (nodal continuity equation):
[0028] Constraint 2 (ring energy equation):
[0029] Constraint 3 (minimum head at node):
[0030] Where: n is the number of pipe segments, m is the number of nodes, q is the number of closed loops; L i , D i are the length (m) and diameter (mm) of the i-th pipe section respectively; C(D i ) is the actual unit price of the i-th pipe section, obtained by comparing the pipe database, D i ∈{D(1), D(2), ..., D(np)}, np is the number of optional commercial pipe diameters; Q j,in , Q j,out are the inflow and outflow flows of node j (l / s); Q j is the flow rate that node j needs to meet (l / s); ΔH k is the head loss (m) of the closed loop k in the annular water supply network; the head loss (H) can be calculated by the following formula: H = ω·Q β ·L / (C W β ·D γ ), where ω, β, and γ are constants; C W is the Hezen-Williams coefficient, obtained by comparing the pipe database; H j , H j,min are the free head (m) and the minimum allowable head (m) at node j, respectively.
[0031] The method of generating the initial tube diameter matrix U by the approximation principle described in step (4) is as follows: First, randomly generate an N×n tube diameter matrix U=(d ij ) N×n Then, the diameter matrix is normalized, that is, the optional commercial diameter value (D i ) replace the elements in the matrix one by one to realize the discretization of the pipe diameter combination, so that each pipe diameter value belongs to the set of optional commercial pipe diameters.
[0032] x p =x min +ξ·(x max -x min ) (8)
[0033] Where: x p is the pth pipe diameter; x min and x maxare the minimum and maximum values of commercial pipe diameters respectively; ξ∈[0,1].
[0034] The specific process of ICS-EPANET in step (4) for automatically iterating and updating the pipe diameter matrix U is as follows: In the tth iteration, first, all matrix rows (x i t ) of the pipeline network investment value (F i t ), through the spiral Lévy flight update method (Formula 1), all x in U i t Generate x i t+1 And calculate F i t+1 , F i t+1 With F i t In comparison, if F i t+1 If it is better, use x i t+1 Replace x i t , otherwise it remains unchanged; then, through the dynamic discovery probability update method (Formula 2), x is again i t Generate new x i t+1 And calculate the new F i t+1 , this new F i t+1 The corresponding Lévy updated F i t+1 In comparison, if the new F i t+1 If it is better, use the new x i t+1 Replaces the x in the Helical Lévy flight update i t+1 , otherwise remain unchanged, complete the tth iteration, and use step (4.4) to enter the next iteration.
[0035] The pipe network information described in step (5) includes: the minimum total cost of investment and construction, the optimal design diameter of each pipe section and its corresponding Hezen-Williams coefficient, the node pressure of each node, the head loss of each pipe section, and the transmission flow of each pipe section.
[0036] The beneficial effects of the present invention are as follows: a method for optimizing the layout of pipe diameters of an urban ring-shaped water supply network is proposed by the present invention. The method automatically generates a discrete nonlinear pipe diameter optimization model that is mutually fed with the local pipe database and aims to minimize the total cost of the pipe network investment by establishing a local pipe database and generating a pipe network system .inp file based on C language; an initial pipe diameter matrix is generated by a random walk method and an approximation principle, and an update method of an adaptive spiral Lévy flight and a dynamic discovery probability is used to realize the update iteration of the pipe diameter matrix that is mutually fed with the local pipe database; EPANET hydraulic judgment is performed through the mutual feedback of the .inp file and the local pipe database to obtain a group of pipe diameter combinations with the minimum total investment and a set of pipe network system information. The method is scientific and reasonable, and effectively solves the problems of multiple isoform constraints, easy to fall into local optimality, pipe diameter discretization, and inaccuracy in determining the pipe diameter by traditional heuristic algorithms in the optimization model of the ring-shaped water supply network. The method is suitable for the pipe network design of complex systems and can globally optimize the investment in pipe network construction. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 The present invention is a flow chart of a method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network.
[0038] Figure 2 This is a layout diagram of the water supply main network according to an embodiment of the present invention (a. node information table; b. optional commercial pipe diameter size prices; c. pipe network layout diagram).
[0039] Figure 3 It is the investment convergence curve and running time curve of the ICS-EPANET algorithm in the embodiment of the present invention. DETAILED DESCRIPTION
[0040] The present invention will be further described below in conjunction with the accompanying drawings:
[0041] like Figure 3 As shown in the figure, a method for optimizing the layout of pipe diameters for urban ring water supply networks is used. The main water supply network of Liyang Central Water Plant is selected as the experimental research object. The network is composed of 22 nodes (node 1 is the source node), 32 pipe sections and 11 rings. The water pressure elevation at the factory is 51.0m, and the minimum allowable water head of each node is 28m.
[0042] like Figure 1 , Figure 2 As shown, the specific implementation steps are as follows:
[0043] (1) Obtain commercial information on local water supply pipes of different pipe materials and establish a local pipe database, including optional commercial pipe inner diameters of common pipe materials such as ductile iron pipes, PE pipes, welded steel pipes, concrete pipes, and their corresponding local commercial prices and Hezen-Williams coefficients.
[0044] (2) Based on the hydraulic analysis engine EPANET, the initial conditions and established layout of the urban ring water supply network system are improved, namely: the established number of each node, the established elevation of each node, the established outflow of each node, the established number of each pipe section, the established length of each pipe section, the initial pipe diameter of each pipe section, the initial Hezen-Williams coefficient of each pipe section, the optional commercial pipe diameter and its corresponding price and Hezen-Williams coefficient, and the .net file of the urban ring water supply network system is output; the network system .net file is read through C language, and attribute categories are added to each data layer of the initial conditions and established layout of the network system, namely: pipe section number, pipe diameter, pipe length, Hezen-Williams coefficient, node number and node demand, and the .inp file is created and the attribute categories and their corresponding network information are written.
[0045] (3) Read the .inp file of the pipe network system in binary mode, automatically extract the data information required for the pipe diameter optimization model in the .inp file according to the attribute category, and automatically generate a pipe diameter optimization model based on the pipe diameter (D) of each pipe segment in combination with the local pipe database. i ) is a discrete decision variable, and the total cost of pipeline network investment is minimized (C min ) is the pipe diameter optimization model of the objective function, and the specific optimization model is as follows:
[0046] Objective function:
[0047] Constraint 1 (nodal continuity equation):
[0048] Constraint 2 (ring energy equation):
[0049] Constraint 3 (minimum head at node):
[0050] Where: n is the number of pipe segments, m is the number of nodes, q is the number of closed loops; L i , D i are the length (m) and diameter (mm) of the i-th pipe section respectively; C(D i ) is the actual unit price of the i-th pipe section, obtained by comparing the pipe database, D i ∈{D(1), D(2), ..., D(np)}, np is the number of optional commercial pipe diameters; Q j,in , Q j,out are the inflow and outflow flows of node j (l / s); Q j is the flow rate that node j needs to meet (l / s); ΔH k is the head loss (m) of the closed loop k in the annular water supply network; the head loss (H) can be calculated by the following formula: Among them, ω, β, γ are constants; C W is the Hezen-Williams coefficient, obtained by comparing the pipe database; H j , H j,min are the free head (m) and the minimum allowable head (m) at node j, respectively.
[0051] (4) The pipe diameter matrix U of the water supply network system is deeply coupled with the improved cuckoo algorithm program (ICS-EPANET) through EPANET, and the pipe diameter matrix U of the pipe network system is automatically iterated and optimized by the mutual feedback of information between the local pipe database, so as to determine the pipe diameter combination with the lowest investment value of the pipe network construction. The specific steps of the ICS-EPANET algorithm are as follows:
[0052] Step 4.1, initialize the ICS-EPANET algorithm parameters: set the number of rows (N) of the pipe diameter matrix U, the upper limit of the probability of discovery (Pa max ), lower limit of discovery probability (Pa min ), maximum step size control factor (α max ) and the maximum number of iterations (t max );
[0053] Step 4.2, t = 1, generate the initial N × n diameter matrix U = (d ij ) N×n The specific steps are as follows: First, randomly generate an N×n pipe diameter matrix U=(d ij ) N×n Then, the diameter matrix is normalized, that is, the optional commercial diameter value (D i ) replace the elements in the matrix one by one to realize the discretization of the pipe diameter combination, so that each pipe diameter value belongs to the set of optional commercial pipe diameters.
[0054] x p =x min +ξ(x max -x min ) (8)
[0055] Where: x p is the pth pipe diameter; x min and x max are the minimum and maximum values of commercial pipe diameters, respectively; ξ∈[0,1];
[0056] Step 4.3, in the tth iteration, ICS-EPANET is used to automatically iterate and update U, and the iteration formula is as follows:
[0057] Updated method for adaptive spiral Lévy flight:
[0058]
[0059] Dynamic Pa update method:
[0060] x i t+1 =x i t +γ·H(Pa max -(Pa max -Pa min )t / t max -ε)·(x j t -x k t ) (2)
[0061] Where: x i t 、x i t+1 are the i-th matrix row in the t-th generation and t+1-th generation pipe diameter matrix U respectively; α is the step size control factor, calculated using formula (3); Lévy(λ) is the random search path matrix, which obeys the Lévy distribution: Lévy(λ)=μ·Φ / |ν| 1 / β , μ and ν follow the standard normal distribution, Φ = [Γ(1+β)·sin(π·β / 2)] / [Γ(1+β)·β·2 (β-1) / 2 / 2], Γ(1+β) is the Γ function, that is β is a constant, β∈[0,2]; x best t is the optimal pipe diameter combination of the tth generation, D t For x i t to x best t The distance, D t =|x best t -x i t |, b is the spiral constant, p and l are random numbers between [0,1], H is the Heaviside function, ε is a random number that follows the generalized Pareto distribution, and ε=(0.83+0.139·x) -6 , x, γ are random numbers that obey uniform distribution, γ∈[0,1], x j t 、x k t are the two random matrix rows in the t-th generation pipe diameter matrix U;
[0062]
[0063] The specific process of ICS-EPANET's automatic iterative update of the pipe diameter matrix U is as follows: In the tth iteration, first, all matrix rows (x i t ) of the pipeline network investment value (F i t ), through the spiral Lévy flight update method (Formula 1), all x in U i t Generate x i t+1 And calculate F i t+1 , F i t+1 With F i t In comparison, if F i t+1 If it is better, use x i t+1 Replace x i t , otherwise it remains unchanged; then, through the dynamic Pa update method (Formula 2), x is again i t Generate new x i t+1 And calculate the new F i t+1 , this new F i t +1 The corresponding Lévy updated F i t+1 In comparison, if the new F i t+1 If it is better, use the new x i t+1 Replaces the x in the Helical Lévy flight update i t+1 , otherwise it remains unchanged, completing the tth iteration and proceeding to the next iteration using step 4.4;
[0064] Step 4.4, the optimized pipe diameter matrix U is discretely valued through the approximation principle of step 2.2, and each matrix row of the new pipe diameter matrix U covers the "pipe diameter" attribute category in the .inp file, and feeds back with the local pipe database information, adjusts the Hezen-Williams coefficient in the .inp file, performs EPANET hydraulic judgment, and discards the matrix rows that do not meet the constraints; then the new matrix rows are supplemented through the random walk method and approximation principle of step 2.2, and the new matrix rows are again subjected to EPANET hydraulic judgment through the .inp file until all matrix rows of the pipe diameter matrix U meet the constraints, and the pipeline network construction investment values of all matrix rows in the pipe diameter matrix U are calculated and sorted, t=t+1;
[0065] Step 4.5, t>t max The iteration is terminated and the optimal pipe diameter matrix row is obtained.
[0066] (5) After the ICS-EPANET optimization of the pipe diameter of the pipe network system is completed, all attribute categories and their corresponding pipe network information in the .inp file corresponding to the row of the optimal pipe diameter matrix are output, including: the minimum total investment and construction cost, the optimal design pipe diameter of each pipe section and its corresponding Hazen-Williams coefficient, the node pressure of each node, the head loss of each pipe section, and the transmission flow of each pipe section.
[0067] like Figure 3 The investment average value is the average value of 100 repeated calculations in each iteration; the minimum investment value is the minimum value among 100 times; the average operation time is the average value of 100 operation times; Table 1 shows the actual pipe diameter, actual investment, optimized optimal pipe diameter and minimum investment of the water supply main network in the embodiment of the present invention; Table 2 shows the actual node pressure and optimized node pressure of the water supply main network in the embodiment of the present invention. Figure 3 As shown, after the pipe network matrix of the embodiment is optimized by the method, the optimal pipe diameter matrix row, the optimal pipe diameter and investment of each pipe section, the water head of each node and the minimum pipe network construction investment are determined.
[0068] Table 1
[0069]
[0070] Note: Pipeline segment investment difference = cost of using the actual pipe diameter for the i-th pipe segment - cost of using the optimized pipe diameter for the i-th pipe segment. Investment difference ratio = Pipe segment investment difference ÷ (total investment of using the actual pipe diameter for the pipe network system - total investment of using the optimized pipe diameter for the pipe network system) × 100%, of which the total investment of using the actual pipe diameter for the pipe network system is 7.41 × 10 7 Yuan, the total investment of optimizing pipe diameter is 6.48×10 7 Yuan.
[0071] Table 2
[0072]
Claims
1. A method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network, characterized in that: The following steps are involved: (1) Obtain commercial information on local water supply pipes of different pipe materials and establish a local pipe database; (2) Based on the hydraulic analysis engine EPANET, the initial conditions and established layout of the urban ring water supply network system are improved, and the .net file of the urban ring water supply network system is output; Read the pipe network system .net file through C language, add attribute categories to the initial conditions of the pipe network system and the data information of the established layout mode, create a .inp file and write the attribute categories and the corresponding pipe network information; (3) Read the .inp file of the pipe network system in binary mode, automatically extract the data information required for the pipe diameter optimization model in the .inp file according to the attribute category, and automatically generate a pipe diameter optimization model that interacts with the local pipe database in combination with the local pipe database; (4) The pipe diameter matrix U of the water supply network system is deeply coupled with the improved cuckoo algorithm program ICS-EPANET through EPANET, which completes the automatic iterative optimization of the pipe diameter matrix U of the pipe network system and the information feedback of the local pipe database, and determines the pipe diameter combination with the lowest investment value of the pipe network construction. The specific ICS-EPANET algorithm steps are as follows: (4.1) Initialize ICS-EPANET algorithm parameters: set the number of rows N of the pipe diameter matrix U and the upper limit of the discovery probability Pa max , the lower limit of the probability of discovery Pa min , maximum step length control factor α max and the maximum number of iterations t max ; (4.2) t = 1, the initial N × n diameter matrix U = (d ij ) N×n ; (4.3) In the tth iteration, ICS-EPANET is used to adaptively iterate U, and its iterative formula is as follows: Updated method for adaptive spiral Lévy flight: Dynamic discovery probability update method: x i t+1 =x i t +γ·H(Pa max -(Pa max -Pa min )t / t max -ε)·(x j t -x k t ) (2) Where: are the i-th matrix row in the t-th and t+1-th generation pipe diameter matrices U respectively; α is the step size control factor, calculated using formula (3); Lévy(λ) is the random search path matrix, which obeys the Lévy distribution: Lévy(λ)=μ·Φ / |ν| 1 / η , μ and ν follow the standard normal distribution, Φ = [Γ(1+η)·sin(π·η / 2)] / [Γ(1+η)·η·2 (η-1) / 2 / 2], Γ(1+η) is the Γ function, that is η is a constant, η∈[0,2]; x best t is the optimal pipe diameter combination of the tth generation, D t for to x best t The distance, D t =|x best t -x i t |, b is the spiral constant, p and l are random numbers between [0,1], H is the Heaviside function, ε is a random number that follows the generalized Pareto distribution, and ε=(0.83+0.139·x) -6 , x, γ are random numbers that obey uniform distribution, γ∈[0,1], x j t 、x k t are the two random matrix rows in the t-th generation pipe diameter matrix U; (4.4) The optimized pipe diameter matrix U is discretely valued through the approximation principle of step (4.2), and each matrix row of the new pipe diameter matrix U covers the "pipe diameter" attribute category in the .inp file, and is fed back with the local pipe database information, and the Hezen-Williams coefficient in the .inp file is adjusted to perform EPANET hydraulic judgment, and the matrix rows that do not meet the constraints are discarded; then the new matrix rows are supplemented through the random walk method and approximation principle of step (4.2), and the new matrix rows are again subjected to EPANET hydraulic judgment through the .inp file until all matrix rows of the pipe diameter matrix U meet the constraints, and the pipeline network construction investment values of all matrix rows in the pipe diameter matrix U are calculated and sorted, t = t + 1; (4.5)t>t max Terminate the iteration and determine the optimal pipe diameter matrix row; (5) After the ICS-EPANET optimization of the pipe diameter of the pipe network system is completed, all attribute categories and their corresponding pipe network information in the .inp file corresponding to the row of the optimal pipe diameter matrix are output.
2. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The pipe database described in step (1) includes: the inner diameters of ductile iron pipes, PE pipes, welded steel pipes, and concrete pipes, and their corresponding commercial prices and Hezen-Williams coefficients.
3. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The initial conditions and established layout of the urban ring water supply network system described in step (2) include: the established number of each node, the established elevation of each node, the established outflow of each node, the established number of each pipe section, the established length of each pipe section, the initial pipe diameter of each pipe section, the initial Hezen-Williams coefficient of each pipe section, the optional commercial pipe diameters and their corresponding prices and Hezen-Williams coefficients.
4. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The attribute categories of the data information described in step (2) include: pipe segment number, pipe diameter, pipe length, Hazen-Williams coefficient, node number and node demand.
5. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The optimization model described in step (3) is based on the diameter D of each pipe segment. i is a discrete decision variable, with the minimum total cost of pipeline network investment C min As the goal, the node continuity equation, the ring energy equation and the node minimum water head are used as constraints; and the optimization model D i Form information feedback with the local pipe database, and i Automatically adjust the actual unit price and Hezen-Williams coefficient of the pipeline in the objective function and constraint conditions. The specific optimization model is as follows: Objective function: Constraint 1 node continuity equation: Constrained 2-ring energy equation: Constrain 3-node minimum water head: Where: n is the number of pipe segments, m is the number of nodes, q is the number of closed loops; L i , D i are the length m and diameter mm of the i-th pipe section respectively; C(D i ) is the actual unit price of the i-th pipe section, obtained by comparing the pipe database, D i ∈{D(1), D(2), ..., D(np)}, np is the number of optional commercial pipe diameters; Q j,in , Q j,out are the inflow and outflow flows of node j, l / s; Q j is the flow rate that node j needs to meet, l / s; ΔH k is the head loss m of the closed loop k in the annular water supply network; where the head loss ΔH k It can be calculated by the following formula: Among them, ω, β, γ are constants; C W is the Hezen-Williams coefficient, obtained by comparing the pipe database; H j , H j,min are the free water head m and the minimum allowable water head m at node j respectively.
6. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The method of generating the initial tube diameter matrix U by random walk and approximation principle described in step (4.2) is: first, randomly generate an N×n tube diameter matrix U=(d ij ) N×n Then, the diameter matrix is normalized, that is, the optional commercial diameter value D that approximates the above diameter combination is used. i Replace the elements in the matrix one by one to achieve discretization of pipe diameter combinations, so that each pipe diameter value belongs to the set of optional commercial pipe diameters; x p =x min +ξ·(x max -x min ) (8) Where: x p is the pth pipe diameter; x min and x max are the minimum and maximum values of commercial pipe diameters respectively; ξ∈[0,1].
7. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The specific process of ICS-EPANET's adaptive iterative update of the pipe diameter matrix U described in step (4.3) is as follows: In the tth iteration, first, all matrix rows x in U are calculated and sorted. i t The investment value of the pipeline network is F i t , through the spiral Lévy flight update method (Formula 1), all x in U i t Generate x i t+1 And calculate F i t+1 , F i t+1 With F i t In comparison, if F i t+1 If it is better, use x i t+1 Replace x i t , otherwise unchanged; Then, by dynamically discovering the probability update method (Formula 2), x i t Generate new x i t+1 And calculate the new F i t+1 , this new F i t+1 The corresponding Lévy updated F i t+1 In comparison, if the new F i t+1 If it is better, use the new x i t+1 Replaces the x in the Helical Lévy flight update i t+1 , otherwise remain unchanged, complete the tth iteration, and use step (4.4) to enter the next iteration.
8. The method for optimizing the arrangement of pipe diameters of a ring-shaped urban water supply network according to claim 1, characterized in that: The pipe network information described in step (5) includes: the minimum total cost of investment and construction, the optimal design diameter of each pipe section and its corresponding Hezen-Williams coefficient, the node pressure of each node, the head loss of each pipe section, and the transmission flow of each pipe section.
Citation Information
Patent Citations
Green building group water supply pipe network optimization method
CN105956705A
Spraying pipe network and layout optimization method thereof
CN112231868A
System and method to optimize operation of a water network
WO2013026731A1
