Optimal dispatching method of pumping stations based on virtual water

Abstract

The present invention discloses a pump station optimization scheduling method based on virtual water in the field of reservoir scheduling, comprising the following steps: 1) obtaining pump station-water diversion area information; 2) constructing a pump station-water diversion area system database; 3) establishing a pump station optimization scheduling model based on virtual water: using C++ to connect to the database, reading the data information of each field of the pump station-water diversion area system in step 2), taking the maximum economic benefit of daily pump station water diversion as the objective function, taking into account the influence of electricity price fluctuations and tidal river sections on the pump station head, and satisfying the constraints of product import, pump station water diversion, water diversion capacity and non-negative decision variables, and establishing a pump station optimization scheduling model based on virtual water that is mutually fed back with the database; 4) for the model, a combined algorithm of 0-1 integer programming and whale algorithm is proposed to solve the model, and the present invention realizes the optimal configuration of regional water resources and achieves the maximum economic benefit during the scheduling period.

Description

Technical Field

[0001] The present invention relates to the technical field of pump station optimization scheduling, and in particular to a pump station optimization scheduling method. Background Art

[0002] Pump station optimization scheduling refers to the use of pump stations to adjust water to meet the water demand of the region in order to maximize economic benefits. Pump station optimization scheduling is a combinatorial optimization problem with strong constraints, nonlinearity, and multi-stage characteristics. Its core is: (1) Constructing a reasonable pump station optimization scheduling model. At present, pump station optimization scheduling mainly focuses on improving the water diversion capacity and operating efficiency of pump stations, reducing the operating costs of pump stations, etc. When constructing the model, it only takes the physical water perspective into consideration, and does not consider the circulation of virtual water (water condensed in goods or services) in the region, which is not conducive to achieving the optimal allocation of water resources. (2) Reduce operating costs. During the actual operation of the pump station, its operating cost will be affected by many factors, such as: electricity prices in different time periods (d i ) changes and the tidal river section's impact on the pump station head (H i ) (3) Choose an appropriate solution method. The methods for solving optimization scheduling problems generally include linear programming, nonlinear programming, dynamic programming and swarm intelligence optimization algorithms, but these methods also inevitably have some problems: ① Linear programming and nonlinear programming have problems such as slow solution speed and low efficiency; ② Dynamic programming has a large amount of calculation and is prone to "dimensionality disaster"; ③ Swarm intelligence optimization algorithms are prone to problems such as low solution accuracy and falling into local optimality. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention provides a method for optimizing the scheduling of pump stations based on virtual water. On the basis of the traditional optimization scheduling of pump stations, not only the circulation of virtual water is considered, but also the actual operation of electricity price d i And lift H i Based on the changes in the regional water resources, a regional virtual water-pumping station optimization scheduling model was constructed; and based on the current solution method, an improved whale optimization algorithm (ZOWOA) coupled with 0-1 integer programming was proposed, thereby achieving the optimal configuration of regional water resources and obtaining the maximum economic benefit during the scheduling period.

[0004] The object of the present invention is achieved by: a method for optimizing the scheduling of a pump station based on virtual water, comprising the following steps:

[0005] Step 1) Obtaining pump station-water diversion area information: Obtaining scheduling parameter information, pump station unit information and water diversion area crop information, and completing the conversion of crop yield and virtual water according to the formula to obtain virtual water information;

[0006] Step 2) Building a pump station-water diversion area system database: Building a pump station-water diversion area system database, and adding the field information of step 1) to the database, and then batch importing the corresponding pump station and water diversion area information data, and updating the data information in real time;

[0007] Step 3) Establish a pump station optimization scheduling model based on virtual water: Use C++ to connect to the database, read the data information of each field of the pump station-water diversion area system in step 2), and use the average daily water diversion volume Y of crops in the irrigation area as the input. ty and the average daily turnover of crops IMP ty The optimal start-up time k of the pumping station in a day is taken as the decision variable, the daily economic benefit of pumping station water diversion max F is taken as the objective function, and the constraints of product import, pumping station water diversion, water diversion capacity and non-negative decision variables are met to generate a virtual water-based pumping station optimization scheduling model that interacts with the database; the specific optimization model is as follows:

[0008] Objective function:

[0009]

[0010] Constraint 1 Product import constraints:

[0011] Constraint 2 Pump station water diversion constraints:

[0012] Constraint 3: Water diversion capacity constraint:

[0013] Constraint 4 variable non-negative constraint: Y ty ≥0,IMP ty ≥0 (5)

[0014] Where: ty is the crop type, ty∈[1,N]; INC ty The income from the pumping station to divert water to crops ty, Y ty is the amount of water that the pumping station diverts to crops ty in one day, EXP ty Cost of pumping stations to bring water to crops; IC ty The income from importing agricultural products, EP ty is the cost of imported agricultural products, IMP ty is the mass of imported agricultural products; p is the medium coefficient, Q is the design flow rate, Δt is the interval time, and its value is 24 / T, where T is the number of time periods, d i is the electricity cost for the i-th hour, k i is the diversion coefficient at the i-th hour, k i =1, water is diverted, k i = 0, no water is introduced, H iis the water lifting head in the ith hour, which is affected by the tide, η is the rated working efficiency of the pumping station, β is the pumping energy coefficient, and d is the unit price of electricity; PRD ty is the local average daily production of agricultural product ty, TD ty is the daily average demand of agricultural products ty; BZ is the daily average maximum water supply of the pumping station;

[0015] (4) Model solving method: For the virtual water-based pump station optimization scheduling model constructed in steps 1)-3), the model is automatically iteratively optimized and solved by the combined algorithm ZOWOA of 0-1 integer programming and whale algorithm to obtain the optimal scheduling plan for the pump station.

[0016] As a further limitation of the present invention, the specific steps of solving in step 4) are as follows:

[0017] Step 1 Parameter setting: Set the number of rows and columns of the variable matrix Va to Num = 2N, the current number of iterations t, and the maximum number of iterations T max , upper and lower limits of daily water diversion of pumping stations, and crop demand TD ty , result variable Res, 1×Dim result matrix RES and 1×Dim power-on time matrix PO; result variable Res is used to obtain the value of max F; result matrix RES is used to obtain the Va(i) value corresponding to maxF; power-on time matrix PO is used to obtain the optimal power-on time k corresponding to max F;

[0018] Step 2 Adaptive Diverse Initialization: Generate a variable matrix Va of Num×Dim, and regard each row of the variable matrix Va as a set of candidate solutions. The candidate solutions are represented by Y ty and IMP ty Composition; Update the random solution through Equation 6 or The specific formula is as follows:

[0019]

[0020] Where: i represents the number of rows in the matrix Va, i∈[1,Num], up is the upper limit of the variable, down is the lower limit of the variable, initialization is the initialization factor, ξ is a random number between [0,1], t is the current iteration number, T max is the maximum number of iterations; is the i-th candidate solution randomly generated for t iterations, is the i-th candidate solution generated by t-iteration uniform distribution, τ is a random number that obeys the uniform distribution of [0,1];

[0021] Step 3 Constraint processing and 0-1 integer programming solution process: First, read the data TD tyand BZ, initialize Va(i) according to Step2; secondly, perform constraint judgment according to Formula 2, Formula 4 and Formula 5, delete the matrix rows Va(i) that do not meet the above constraints, and regenerate them through Step 2; then, repeat the above process until all candidate solution matrices Va meet the constraints; finally, establish the linear programming model Formula 12, and solve the linear programming model Formula 12 that meets the constraints of Formula 3 and Formula 13 through 0-1 integer programming, and obtain the optimal startup time k corresponding to Va(i);

[0022] Step 4 Iterative update process: In the tth iteration, first calculate the fitness value and get the optimal solution and the optimal start-up time k; and update the algorithm parameters according to the corresponding formula, including: a, A, C, ω1, ω2 and p; on this basis, according to the p and |A| values, each whale performs an iterative update strategy of encirclement predation, random or spiral bubble; the specific update method is: when p<0.5, |A|≤1, use formula 8 for encirclement predation update; when p<0.5, |A|>1, use formula 7 for random update; when p≥0.5, use formula 9 for spiral bubble update; then, according to the boundary neighborhood update formula 10, the out-of-bounds candidate solution Va(i) is corrected and updated to obtain the candidate solution matrix Va of the t+1 generation; the specific update formula is as follows:

[0023] Random search update method: Va(i) t+1 =Va(rand) t -A·|C·Va(rand) t -Va(i) t | (7)

[0024] Surround and predation update method:

[0025] Spiral bubble update method:

[0026] Boundary neighborhood update method:

[0027] In the formula: Va(i) t is the i-th group of candidate solutions in the t iteration; is the optimal solution in t iterations; Va(rand) tis a set of solutions randomly selected from the t-time iteration matrix Va; A and C are matrix coefficients, expressed as: A = 2a·(r1-1) and C = 2a·r2 respectively; a is the convergence factor, the convergence curve of the coupled fastest curve is expressed as a = a0-K0(1-cosθ), a0 is the starting point of the curve, the value is 2, K0 is the coefficient, θ is the angle between the tangent line at a certain point on the convergence curve and the vertical axis; r1 and r2 are random vectors of [0,1]; the weight change curve of the coupled fastest curve is expressed as ω1 = ω0-0.5K0(1-cosθ), ω2 = 1.2ω0+0.4K0(1-cosθ), ω0 is 1; D′ represents the distance between the i-th candidate solution and the optimal solution in the t-th iteration; b is the spiral shape constant, which takes the value of 1; l is a random number in [-1,1], whose size determines the distance between the candidate solution and the optimal solution, l=-1 is the closest distance, and l=1 is the farthest; U(0,0.1) means that it obeys the uniform distribution between [0,0.1];

[0028] Step 5: Determine the iteration condition: when t≤T max Repeat Step 4 until the iteration is exited, and output the results as a .txt file, and draw a convergence curve according to the Res value of each iteration; the result information includes:

[0029] (1) Virtual water information VW ty ,

[0030] (2) Optimal daily water diversion volume Y ty and the average daily import volume of agricultural products IMP ty ,

[0031] (3) Optimal start-up time k for the pump,

[0032] (4) Maximum economic benefit max F.

[0033] As a further limitation of the present invention, the linear programming model established in Step 3 is as follows:

[0034] Objective function:

[0035] Constraints:

[0036] in, Indicates greater than Y ty / The smallest integer for Q.

[0037] As a further limitation of the present invention, the optimal solution is obtained in Step 4. The specific process is as follows: in t iterations, calculate the fitness value max F according to Va(i) and PO, and compare Res with max F; if max F≤Res, keep Res and RES unchanged; if max F>Res, assign max F to Res, assign Va(i) to RES, and record the startup time k of the linear programming solution to PO until the calculation of the matrix Va is completed. At this time,

[0038] As a further limitation of the present invention, the method for generating the convergence curve of the coupled fastest curve in Step 4 is:

[0039] (1) Determine the range of the convergence factor a from a0 to 0, and take the special solution point as (x0, y0) = (1, a0);

[0040] (2) Solve θ0 according to the brachistochrone curve formula to obtain K0;

[0041] (3) Combine the algorithm process and the curve to obtain each iteration point, that is: θ0 / T max is the step size factor, discretizing θ into T max point θ∈[0,θ0], then the convergence curve of the coupled fastest curve can be expressed as Equation 14; where Equation 15 is the fastest curve formula;

[0042] a=a0-K0(1-cosθ) (14)

[0043]

[0044] In formula (15), x represents the horizontal coordinate, and its positive direction is to the right; y represents the vertical coordinate, and its positive direction is to the downward.

[0045] As a further limitation of the present invention, the specific method of outputting the result as a .txt file in Step 5 is to add a path in the program and create a .txt file, open the .txt file, and add header field information to the file, namely: (1) virtual water information VW ty (2) Optimal daily water diversion volume Y ty and agricultural product imports IMP ty , (3) the optimal start-up time k of the pump and (4) the maximum economic benefit max F; then, read the corresponding data VW1, VW2, Y ty 、IMP ty , the optimal start-up time k and max F of the pump are written into the header position corresponding to the file, and the file is closed after completion.

[0046] As a further limitation of the present invention, the scheduling parameter information in step 1) includes the total number of time periods T, the number of crop types N, the medium coefficient p, the flow energy coefficient β and the unit price of electricity di The pump station unit information includes pump flow Q, head H i and work efficiency η, the average daily water diversion volume BZ of the pump station; among which, the pump station head H i The design includes the following steps:

[0047] (1) Determine the average daily head H;

[0048] (2) Analyze the tidal process and determine the average high tide level, average tide level, average low tide level, tidal range, average high tide time and average low tide time;

[0049] (3) Perform curve fitting.

[0050] As a further limitation of the present invention, the information on crops in the water diversion area in step 1) includes: the cost EXP of diverting water to the crop ty ty , income from diverting water to crops INC ty 、The average daily demand of crops tyTD ty 、Average daily yield information of crops PRD ty , circulation cost EP ty and revenue information IC ty .

[0051] As a further limitation of the present invention, the virtual water information in step 1), i.e., the virtual water volume, is calculated by a conversion formula of agricultural products-virtual water, and the specific formula is as follows:

[0052]

[0053] Among them: ty is the crop type, VW ty The water footprint of agricultural crops, ET ty is the green and blue evapotranspiration during the crop growth period, Y ty is the yield of crop ty.

[0054] As a further limitation of the present invention, in step 2), the field information of the pump station-water diversion area is used to construct a pump station-water diversion area database, and the above field information is added, including the total number of time periods T, the number of crop types N, the medium coefficient p, the flow energy coefficient β and the unit price of electricity d i ; Pump information flow Q, head H i and work efficiency η, the average daily water diversion volume BZ of the pumping station; the cost of diverting water to crops ty EXP ty , income from diverting water to crops INC ty 、The average daily demand of crops tyTD ty 、Average daily yield information of crops PRD ty , circulation cost EP ty, income information IC ty , virtual water information and; After the fields are added, it is d i , H i Add the primary key index with BZ, and import the corresponding data of each field in batches to complete the database construction.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] (1) This method constructs a pump station-water diversion area system database and stores dispatch parameter data, pump station unit information, water diversion area crop data, and virtual water volume calculation data in batches into the database, which facilitates data query, reading, and real-time update, and is easy to realize data sharing among multiple users;

[0057] (2) This method reads the pump station-water diversion area system database information in real time and automatically generates a pump station optimization scheduling model that considers virtual water and interacts with the database, thus achieving the optimal allocation of regional water resources. ty ), annual crop turnover (IMP ty ) and the most appropriate start-up time (k) are used as decision variables, and the maximum economic benefit of reservoir water supply (max F) during the dispatch period is used as the objective function. At the same time, the constraints of product import, pump station water diversion, water diversion capacity and non-negative decision variables must be met;

[0058] (3) The solution to the pump station optimization scheduling model is based on the improved whale optimization algorithm (ZOWOA) coupled with 0-1 integer programming, which improves the convergence accuracy and speed of solving the optimization scheduling problem of complex water conservancy systems. ① This method generates the initial solution through adaptive diverse initialization, and takes boundary neighborhood updates for out-of-bounds candidate solutions to improve the quality of the solution and the convergence speed of the algorithm; ② This method is coupled with 0-1 integer programming to achieve accurate solution of the optimal startup time k; ③ This method improves the algorithm parameters and introduces the fastest curve, which accelerates the process of whale population optimization and improves the optimization accuracy of the algorithm; ④ This method introduces and improves the weight strategy of the particle swarm algorithm and proposes a dual-weight perturbation update strategy, which solves the problem that the algorithm is prone to fall into local optimality in the later stage and further improves the accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0060] Figure 1This is a technical roadmap for a pump station optimization scheduling method based on virtual water in the present invention.

[0061] Figure 2 The generalized tide level diagram of the tidal river section pump station with an average daily head of 17m in the present invention

[0062] Figure 3 It is a profit convergence curve diagram of the ZOWOA algorithm in the embodiment of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] The pump station near the Jinniushan Reservoir in Liuhe District, Nanjing was selected as the experimental research object. The specific known scheduling parameter information, pump station unit information, and water diversion area crop information are shown in Tables 1-4.

[0065] like Figure 1 As shown, the specific implementation steps are as follows:

[0066] (1) Obtaining pump station-water diversion area information: Obtain scheduling parameter information (Tables 1 and 2), pump station unit information (Table 3), and water diversion area crop information (Table 4), and complete the conversion of crop yield and virtual water according to the formula to obtain virtual water information.

[0067] (2) Constructing a pump station-water diversion area system database: Constructing a pump station-water diversion area system database, and adding the field information of step (1) to the database, and then batch importing the corresponding pump station and water diversion area information data, and updating the data information in real time.

[0068] (3) Establishment of a pump station optimization scheduling model based on virtual water: Use C++ to connect to the database, read the data information of each field of the pump station-water diversion area system in step (2), and automatically generate a pump station optimization scheduling model based on virtual water that interacts with the database. This model takes into account the fluctuation of electricity prices in different time periods and the impact of tidal river sections on the pump station head. The average daily water diversion volume of crops in the irrigation area (Y ty ) and the average daily turnover of crops (IMP ty ) and the optimal start-up time of the pumping station in a day (k) are taken as decision variables, the maximum economic benefit of daily pumping station water diversion (max F) is taken as the objective function, and the constraints of product import, pumping station water diversion, water diversion capacity and non-negative decision variables are met; the specific optimization model is as follows:

[0069] Objective function:

[0070]

[0071] Constraint 1 (product import constraint):

[0072] Constraint 2 (pump station water diversion constraint):

[0073] Constraint 3 (water diversion capacity constraint):

[0074] Constraint 4 (variable non-negative constraint) Y ty ≥0,IMP ty ≥0(5)

[0075] Where: ty is the crop type, ty∈[1,N]. ty Income from pumping stations to divert water to crops (yuan / m 3 ), Y ty is the amount of water that the pumping station diverts to crops ty in one day (m 3 ), EXP ty Cost of pumping station to divert water to crops (yuan / m 3 );IC ty is the income from imported agricultural products (yuan / kg), EP ty is the cost of imported agricultural products (yuan / kg), IMP ty For the quality of imported agricultural products (10 4 Kg). p is the medium coefficient, Q is the design flow rate (m 3 / s), Δt is the interval time, and its value is 24 / T, where T is the number of time periods, d i is the electricity cost for the i-th hour, k i is the diversion coefficient at the i-th hour, k i =1, water is diverted, k i = 0, no water is introduced, H i is the water lifting head in the ith hour, affected by the tide, η is the rated working efficiency of the pumping station, β is the pumping energy coefficient, and d is the unit price of electricity. ty is the local average daily production of agricultural product ty (10 4 Kg), TD ty is the daily average demand of agricultural products ty. BZ is the daily average maximum water supply of the pumping station.

[0076] (4) Model solving method: For the virtual water-based pump station optimization scheduling model constructed in steps (1) to (3), the model is automatically iteratively optimized and solved by the whale algorithm (ZOWOA) coupled with 0-1 integer programming. The specific steps of ZOWOA are as follows:

[0077] Step 1 ZOWOA algorithm parameter setting: set the number of rows (Num) and columns (Dim = 2N) of the variable matrix Va, the current number of iterations (t), the maximum number of iterations (T max ), upper and lower limits of daily water diversion of pumping stations, and crop demand (TD ty ), result variable (Res: used to obtain the value of max F), 1×Dim result matrix (RES: used to obtain the Va(i) value corresponding to max F) and 1×Dim power-on time matrix (PO: used to obtain the optimal power-on time k corresponding to max F).

[0078] Step 2 Adaptive Diverse Initialization: Generate a variable matrix Va of Num×Dim, and regard each row of the variable matrix Va as a set of candidate solutions. The candidate solutions are represented by Y ty and IMP ty Composition; Update the random solution through Equation 6 or The specific formula is as follows:

[0079]

[0080] Where: i represents the number of rows in the matrix Va, i∈[1,Num], up is the upper limit of the variable, down is the lower limit of the variable, initialization is the initialization factor, ξ is a random number between [0,1], t is the current iteration number, T max is the maximum number of iterations; is the i-th candidate solution randomly generated for t iterations, is the i-th candidate solution generated by t-iteration uniform distribution, and τ is a random number that follows a uniform distribution in [0,1].

[0081] Step 3 Constraint processing and 0-1 integer programming solution process: First, read the data (TD ty and BZ), initialize Va(i) according to Step2; secondly, perform constraint judgment according to (Equation 2, Equation 4, and Equation 5), delete the matrix rows Va(i) that do not meet the above constraints, and regenerate them through Step2; then, repeat the above process until all candidate solution matrices Va meet the constraints; finally, establish a linear programming model (Equation 12), and solve the linear programming model (Equation 12) that meets the constraints of (Equation 3 and Equation 13) through 0-1 integer programming, and obtain the optimal startup time k corresponding to Va(i).

[0082] Step 4 Iterative update process: In the tth iteration, first calculate the fitness value and get the optimal solution and the optimal start-up time k; and update the algorithm parameters (including: a, A, C, ω1, ω2 and p) according to the corresponding formula. On this basis, according to the values ​​of p and |A|, each whale performs an iterative update strategy of encirclement predation, random or spiral bubble. The specific update method is: when p<0.5, |A|≤1, use (Formula 8) for encirclement predation update; when p<0.5, |A|>1, perform random update according to (Formula 7); when p≥0.5, use (Formula 9) for spiral bubble update. Subsequently, the out-of-bounds candidate solution Va(i) is corrected and updated according to the boundary neighborhood update (Formula 10) to obtain the candidate solution matrix Va of the t+1th generation. The specific update formula is as follows:

[0083] Random search update method: Va(i) t+1 =Va(rand) t -A·|C·Va(rand) t -Va(i) t | (7)

[0084] Surround and predation update method:

[0085] Spiral bubble update method:

[0086] Boundary neighborhood update method:

[0087] In the formula: Va(i) t is the i-th group of candidate solutions in the t iteration; is the optimal solution in t iterations; Va(rand) t is a set of solutions randomly selected from the t-time iteration matrix Va; A and C are matrix coefficients, expressed as: A = 2a·(r1-1) and C = 2a·r2 respectively; a is the convergence factor, the convergence curve of the coupled fastest curve is expressed as a = a0-K0(1-cosθ), a0 is the starting point of the curve, the value is 2, K0 is the coefficient, θ is the angle between the tangent line at a certain point on the convergence curve and the vertical axis; r1 and r2 are random vectors of [0,1]; the weight change curve of the coupled fastest curve is expressed as ω1 = ω0-0.5K0(1-cosθ), ω2 = 1.2ω0+0.4K0(1-cosθ), ω0 is 1; D′ represents the distance between the i-th candidate solution and the optimal solution in t iterations; b is the spiral shape constant, which takes a value of 1; l is a random number in [-1,1], whose size determines the distance between the candidate solution and the optimal solution, l=-1 is the closest, and l=1 is the farthest; U(0,0.1) means that it obeys the uniform distribution between [0,0.1].

[0088] Step 5: Determine the iteration condition: when t≤T maxRepeat Step 4 until the iteration is exited, and output the results as a .txt file, and draw a convergence curve according to the Res value of each iteration, see Figure 3 The result information includes: (1) virtual water information (VW ty ), (2) Optimal daily water diversion (Y ty ) and the average daily import volume of agricultural products (IMP ty ), (3) the optimal start-up time k of the pump and (4) the maximum economic benefit (max F).

[0089] Specifically, the scheduling parameter information described in step (1) includes the total number of time periods (T), the number of crop types (N), the medium coefficient (p), the flow energy coefficient (β) and the unit price of electricity (d i ).

[0090] Specifically, the pump station unit information described in step (1) includes pump information (flow rate Q, head H i and work efficiency η), the average daily water diversion volume BZ of the pump station. Among them, the pump station head H i The design includes the following steps:

[0091] (1) Determine the average daily head H;

[0092] (2) Analyze the tidal process and determine the average high tide level, average tide level, average low tide level, tidal range, average high tide time, and average low tide time;

[0093] (3) Perform curve fitting (see Figure 2 ). Specifically, the crop information of the water diversion area described in step (1) includes: the cost of diverting water to the crop ty (EXP ty ), income from diverting water to crops (INC ty ), the average daily demand of crops ty TD ty , crop daily average yield information (PRD ty ), circulation cost (EP ty ) and revenue information (IC ty ).

[0094] Specifically, the virtual water information in step (1) is the virtual water volume, which is calculated by the agricultural product-virtual water conversion formula; the specific formula is as follows:

[0095]

[0096] Among them: ty is the crop type, VW ty The water footprint of agricultural crops, ET ty is the green and blue evapotranspiration during the crop growth period, Y ty is the yield of crop ty.

[0097] Specifically, the field information of the pump station-water diversion area in step (2) is used to build a pump station-water diversion area database, and the above field information is added, including the total number of time periods (T), the number of crop types (N), the medium coefficient (p), the flow energy coefficient (β) and the unit price of electricity (d i ); Pump information (flow rate Q, head H i and work efficiency η), the average daily water diversion volume BZ of the pumping station; the cost of diverting water to crops ty (EXP ty ), income from diverting water to crops (INC ty ), the average daily demand of crops ty TD ty , crop daily average yield information (PRD ty ), circulation cost (EP ty )、Income Information(IC ty ) and virtual water information. After the field is added, it is d i , H i Add the primary key index with BZ, and import the corresponding data of each field in batches to complete the database construction.

[0098] Specifically, the linear programming model established in Step 3 is as follows

[0099] Objective function:

[0100] Constraints:

[0101] in, Indicates greater than Y ty / The smallest integer for Q.

[0102] Specifically, the optimal solution is obtained in Step 4 The specific process is as follows: in t iterations, the fitness value (max F) is calculated based on Va(i) and PO, and Res is compared with max F. If max F≤Res, Res and RES are kept unchanged; if max F>Res, max F is assigned to Res, Va(i) is assigned to RES, and the startup time k of the linear programming solution is recorded in PO until the calculation of the matrix Va is completed. At this time,

[0103] Specifically, the convergence curve of the coupled fastest curve in Step 4 is generated as follows: (1) determine the range of the convergence factor (a) (a0 to 0), and take the special solution point as (x0, y0) = (1, a0); (2) solve θ0 according to the fastest curve formula to obtain K0; (3) combine the algorithm process and the curve to obtain each iteration point, that is: (θ0 / T max ) is the step size factor, discretizing θ into Tmax points (θ∈[0,θ0]), the convergence curve of the coupled fastest curve can be expressed as Equation 14. Wherein, Equation 15 is the fastest curve formula.

[0104] a=a0-K0(1-cosθ) (14)

[0105]

[0106] In Formula 15, x represents the horizontal coordinate, and its positive direction is to the right; y represents the vertical coordinate, and its positive direction is to the downward.

[0107] Specifically, the specific method of outputting the results to a .txt file in Step 5 is to add a path in the program and create a .txt file, open the .txt file, and add header field information to the file, namely: (1) virtual water information (VW ty ), (2) Optimal daily water diversion (Y ty ) and agricultural product imports (IMP ty ), (3) the optimal start-up time of the pump (k) and (4) the maximum economic benefit (max F). Then, read the corresponding data (VW1, VW2, Y ty 、IMP ty , the optimal start-up time k of the pump and max F), write them to the header position corresponding to the file. Close the file after completion.

[0108] like Figure 2 As shown in the figure, according to the pump station head and the local high tide information, the average daily head H is determined, and the tidal level engineering is analyzed to further determine the average high tide level, average tide level, average low tide level, tidal range, average high tide time and average low tide time. Finally, curve fitting is performed to design the tidal level diagram of the pump station with an average daily head of 17m.

[0109] like Figure 3 As shown, according to the iterative process, the Res (max F) obtained in each iteration is plotted into a curve, and the economic benefit convergence curve diagram in the embodiment of the present invention can be obtained.

[0110] Table 1 Scheduling parameter information

[0111]

[0112] Table 2 1d time period division and peak and valley electricity prices in each period

[0113]

[0114] Note: Peak and valley time-of-use electricity prices announced by the Jiangsu Provincial Price Bureau in July 2008

[0115] Table 3 Pump station unit information

[0116]

[0117] Table 4 Crop information in the water diversion area

[0118]

[0119] Table 5 Output results of virtual water information, optimal water diversion volume and agricultural product import volume, optimal startup time, and maximum economic benefit

[0120]

[0121] In the table, Y ty : represents the average daily water diversion from the pumping station to the crop ty (ty∈[1,2]); IMP ty : represents the average daily circulation of crop products ty; k i : represents the diversion coefficient in hour i.

[0122] Table 5 is the output result information of the embodiment of the present invention, including virtual water information (VW ty ), average daily water diversion for crops (Y ty ) and the average daily import volume of agricultural products (IMP ty ), the most appropriate start-up time (k) and the maximum profit (max F) of the pumping station, which is the optimal dispatching plan for the pumping station.

[0123] The above embodiments are only used to help understand the method and core idea of ​​the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the scheduling of pump stations based on virtual water, characterized in that: The following steps are involved: Step 1) Obtaining pump station-water diversion area information: Obtaining scheduling parameter information, pump station unit information and water diversion area crop information, and completing the conversion of crop yield and virtual water according to the formula to obtain virtual water information; Crop information in the water diversion area, including: the cost of diverting water to the crop ty EXP ty , income from diverting water to crops INC ty 、The average daily demand of crops tyTD ty 、Local average daily production of crops ty PRD ty , the cost of crops ty EP ty and income from crops ty ; Virtual water information, i.e. virtual water volume, is calculated through the crop-virtual water conversion formula. The specific formula is as follows: ； Among them: ty is the crop type, VW ty Virtual water information for crops, ET ty is the green and blue evapotranspiration during the crop growth period, Y ty It is the amount of water that the pumping station diverts to the crops in one day; Step 2) Build a pump station-water diversion area system database: Build a pump station-water diversion area system database, and add the field information of step 1) in the database, then import the corresponding pump station and water diversion area information data in batches, and update the data information in real time; Step 3) Establish a pump station optimization scheduling model based on virtual water: Use C++ to connect to the database and read the data information of each field of the pump station-water diversion area system in step 2) to calculate the water diversion volume of crops ty by the pump station in the irrigation area in one day. and the quality of crops and the water diversion coefficient k of the pumping station at each hour of the day i As the decision variable, the daily economic benefit of pumping station water diversion max F is taken as the objective function, and the constraints of crop import, pumping station water diversion, water diversion capacity and non-negative decision variables are met to generate a virtual water-based pumping station optimization scheduling model that feeds back to the database; the specific optimization model is as follows: Objective function: ； Constraint 1: Agricultural crop import constraints: ; Constraint 2 Pump station water diversion constraints: ; Constraint 3: Water diversion capacity constraint: ; Constraint 4 variable non-negative constraint: ; Where: ty refers to a specific crop, ty∈[1,N]; p is the medium coefficient, Q is the design flow rate, is the interval time, and its value is 24 / T, where T is the number of time periods, d i is the electricity cost for the i-th hour, k i is the diversion coefficient at the i-th hour, k i =1, water is diverted, k i =0 means no water diversion, H i is the water lifting head in the ith hour, which is affected by the tide, η is the rated working efficiency of the pumping station, β is the pumping energy coefficient; BZ is the average maximum daily water supply of the pumping station; Step 4) Model solving method: For the virtual water-based pump station optimization scheduling model constructed in steps 1)-3), the model is automatically iteratively optimized and solved through the combined algorithm ZOWOA of 0-1 integer programming and whale algorithm to obtain the optimal scheduling plan for the pump station.

2. The method for optimizing the scheduling of pump stations based on virtual water according to claim 1, characterized in that: The specific steps for solving step 4) are as follows: Step 1 Parameter setting: Set the number of rows Num, the number of columns Dim, the current number of iterations t, and the maximum number of iterations T of the variable matrix Va max , upper and lower limits of the average daily water diversion of the pumping station, and the average daily demand of crops ty TD ty , result variable Res, 1×Dim result matrix RES and 1×Dim water diversion coefficient matrix PO for each hour of the pumping station in one day; The result variable Res is used to obtain the daily economic benefit of water diversion of the pumping station max F; the result matrix RES is used to obtain the j-th group of candidate solutions Va(j) corresponding to max F; the water diversion coefficient matrix PO of the pumping station at each hour in a day is used to obtain the optimal water diversion coefficient k of the pumping station at each hour in a day corresponding to max F i ; Step 2 Adaptive Diverse Initialization: Generate a variable matrix Va of Num×Dim, where Dim=2N. Each row of the variable matrix Va is considered as a set of candidate solutions. The candidate solutions are and Composition; Update the random solution through formula (6) or , the specific formula is as follows: ； Where: j represents the number of rows in the matrix Va, j∈[1,Num]; up is the upper limit of the variable, referring to Y ty and IMP ty The maximum value of Y; down is the lower limit of the variable, referring to Y ty and IMP ty The minimum value of; initialization is the initialization factor; is a random number between [0,1]; t is the current iteration number; T max is the maximum number of iterations; The jth group of candidate solutions is randomly generated for t iterations; is the jth group of candidate solutions generated by uniform distribution of t iterations; τ is a random number between [0,1] that obeys uniform distribution and can be directly generated by computer functions; Step 3 Constraint processing and 0-1 integer programming solution process: First, read the data TD ty and BZ, initialize Va(j) according to Step2; secondly, perform constraint judgment according to equations (2), (4) and (5), delete the matrix rows Va(j) that do not meet the above constraints, and regenerate them through Step2; then, repeat the above process until all candidate solution matrices Va meet the constraints; finally, establish a linear programming model, that is, equation (12), and solve the linear programming model (12) that meets the constraints of equations (3) and (13) through 0-1 integer programming, and obtain the optimal water diversion coefficient k of the pumping station at each hour in one day corresponding to Va(j) i ; The established linear programming model is as follows Objective function: ; Constraints: ; in, Indicates greater than Y ty / The smallest integer of Q; Step 4 Iterative update process: In the tth iteration, first calculate the fitness value max F and get the optimal solution and the optimal water diversion coefficient k of the pumping station at each hour in a day i ; and update the algorithm parameters according to the corresponding formula, including: a, A, C, ω1, ω2 and p; on this basis, according to the p and |A| values, each whale performs an iterative update strategy of encirclement predation, random or spiral bubble; the specific update method is: when p<0.5, |A|≤1, use formula (8) for encirclement predation update; when p<0.5, |A|>1, use formula (7) for random update; when p≥0.5, use formula (9) for spiral bubble update; then, according to the boundary neighborhood update formula (10), the out-of-bounds candidate solution Va(j) is corrected and updated to obtain the candidate solution matrix Va of the t+1th generation; the specific update formula is as follows: Random search update method: ; Surround and predation update method: ; Spiral bubble update method: ; Boundary neighborhood update method: ; Where: is the jth group of candidate solutions in the t iteration; is the optimal solution in t iterations; is a set of solutions randomly selected from the t-time iteration matrix Va; A and C are matrix coefficients, expressed as: A=2a·(r1-1) and C=2a·r2; a is the convergence factor, and the convergence curve of the coupled fastest curve is expressed as , a0 is the starting point of the curve, with a value of 2, K0 is the coefficient, θ is the angle between the tangent line at a point on the convergence curve and the vertical axis; r1 and r2 are random vectors of [0,1]; the weight change curve of the coupled fastest curve is expressed as , , ω0 takes the value of 1; , ω1 and ω2 are weight coefficients in [0,1], Indicates the distance between the jth group of candidate solutions and the optimal solution in t iterations; b is the spiral shape constant, which takes a value of 1; l is a random number in [-1, 1], whose size determines the distance between the candidate solution and the optimal solution, l=-1 is the closest, and l=1 is the farthest; U(0,0.1) means that it obeys the uniform distribution between [0,0.1]; Step 5 Determine the iteration condition: when t≤T max Repeat Step 4 until the iteration is exited, and output the results as a .txt file, and draw a convergence curve according to the Res value of each iteration; the result information includes: (1) Virtual water information VW ty , (2) The amount of water Y that the pumping station diverts to crops ty in one day ty and the average daily crop quality IMP ty , (3) Optimal pumping unit: water diversion coefficient k for each hour of the pumping station in a day i , (4) Maximum economic benefit max F.

3. The method for optimizing the scheduling of pump stations based on virtual water according to claim 2, characterized in that: Get the optimal solution in Step 4 The specific process is as follows: in t iterations, calculate the fitness value max F according to Va(j) and PO, and compare Res with max F; if max F≤Res, keep Res and RES unchanged; if max F>Res, assign max F to Res and assign Va(j) to RES, and record the water diversion coefficient k of the pumping station at each hour in one day by linear programming solution. i to PO until the calculation of the matrix Va is completed. =Res.

4. The method for optimizing the scheduling of pump stations based on virtual water according to claim 2, characterized in that: The method for generating the convergence curve of the coupled fastest curve in Step 4 is: (1) Determine the range of the convergence factor a from a0 to 0, and take the special solution point as (x0, y0) = (1, a0); (2) Solve θ0 according to the brachistochrone curve formula to obtain K0; (3) Combining the algorithm process and the curve, we can get each iteration point, that is, θ0 / T max is the step size factor, discretizing θ into T max point θ∈[0,θ0], then the convergence curve of the coupled fastest curve can be expressed as formula (14); where formula (15) is the fastest curve formula; ； ； In formula (15), x represents the horizontal coordinate, and its positive direction is to the right; y represents the vertical coordinate, and its positive direction is downward; (x0, y0) is the coordinate of the starting point of the curve, θ0 is the angle between any point on the curve and the Y axis; K0 is the coefficient of the fastest curve.

5. The method for optimizing the scheduling of pump stations based on virtual water according to claim 2, characterized in that: The specific method of outputting the results to a .txt file in Step 5 is to add a path in the program and create a .txt file, open the .txt file, and add header field information to the file, namely: (1) Virtual water information VW ty (2) The amount of water Y that the pumping station diverts to crops ty in one day ty and the average daily crop quality IMP ty (3) Optimal pumping unit: the water diversion coefficient k of the pumping station at each hour of the day i and (4) maximum economic benefit max F; then, read the corresponding data VW respectively ty , Y ty 、IMP ty , the optimal pumping station water diversion coefficient k for each hour in a day i and max F, write to the header position corresponding to the file, and close the file after completion.

6. The method for optimizing the scheduling of pump stations based on virtual water according to claim 1, characterized in that: The scheduling parameter information described in step 1) includes the total number of time periods T, the number of crop types N, the medium coefficient p, the pumping energy coefficient β and the electricity fee d at the i-th hour i The pump station unit information includes the pump design flow Q, head H i , the rated working efficiency η of the pumping station and the average maximum daily water supply of the pumping station BZ; among which, the pumping station head H i The design includes the following steps: (1) Determine the average daily head H; (2) Analyze the tidal process and determine the average high tide level, average tide level, average low tide level, tidal range, average high tide time and average low tide time; (3) Perform curve fitting.

7. The method for optimizing the scheduling of pump stations based on virtual water according to claim 1, characterized in that: In step 2), the pump station-water diversion area field information is constructed to build a pump station-water diversion area database and add the above field information, including the total number of time periods T, the number of crop types N, the medium coefficient p, the pumping energy coefficient β and the electricity fee d for the i-th hour i ; Pump design flow Q, head H i and the rated working efficiency η of the pumping station, the average maximum daily water supply of the pumping station BZ; the cost of diverting water to crops ty EXP ty , income from diverting water to crops INC ty 、The average daily demand of crops tyTD ty 、Local average daily production of crops ty PRD ty , the cost of crops ty EP ty , income from crops ty 、Virtual Water Information VW ty ; After the field is added, it is d i , H i Add the primary key index with BZ, and import the corresponding data of each field in batches to complete the database construction.

Images