CN116562537A - Floodgate pump group flood control and drainage real-time optimal scheduling method, system and storage medium - Google Patents
Floodgate pump group flood control and drainage real-time optimal scheduling method, system and storage medium Download PDFInfo
- Publication number
- CN116562537A CN116562537A CN202310283301.9A CN202310283301A CN116562537A CN 116562537 A CN116562537 A CN 116562537A CN 202310283301 A CN202310283301 A CN 202310283301A CN 116562537 A CN116562537 A CN 116562537A
- Authority
- CN
- China
- Prior art keywords
- scheduling
- drainage
- gate
- pump group
- flood control
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 69
- 238000003860 storage Methods 0.000 title claims abstract description 17
- 238000004364 calculation method Methods 0.000 claims abstract description 17
- 238000013461 design Methods 0.000 claims abstract description 11
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 80
- 238000012360 testing method Methods 0.000 claims description 40
- 238000011156 evaluation Methods 0.000 claims description 36
- 238000012549 training Methods 0.000 claims description 32
- 230000000694 effects Effects 0.000 claims description 28
- 230000006870 function Effects 0.000 claims description 19
- 238000005457 optimization Methods 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000010206 sensitivity analysis Methods 0.000 claims description 12
- 238000012216 screening Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 7
- 238000001363 water suppression through gradient tailored excitation Methods 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000011144 upstream manufacturing Methods 0.000 claims description 6
- 229910052799 carbon Inorganic materials 0.000 claims description 5
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 4
- 230000002265 prevention Effects 0.000 claims description 4
- 230000001133 acceleration Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 230000004907 flux Effects 0.000 claims description 3
- 230000005484 gravity Effects 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 238000004088 simulation Methods 0.000 abstract description 5
- 239000004576 sand Substances 0.000 description 10
- 238000005086 pumping Methods 0.000 description 4
- 238000009472 formulation Methods 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000000988 bone and bone Anatomy 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Quality & Reliability (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Operations Research (AREA)
- Pure & Applied Mathematics (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a floodgate pump group flood control and drainage real-time optimal scheduling method, a floodgate pump group flood control and drainage system and a storage medium. The method constructs a brake pump scheduling model based on LSTM, and optimizes a scheduling scheme by using the brake pump scheduling model of LSTM. The LSTM model is a special RNN (RecurrentNeuralNetwork) model, and the special structural design of the LSTM model enables the LSTM model to consider the influence of the historical time information on the current scheduling scheme. The LSTM model is trained by adopting an optimized scheduling scheme generated by a scheduling model based on hydrodynamic force calculation, so that the LSTM model can learn to obtain the relation between the optimized scheduling scheme and each decision variable. The trained LSTM model is used for generating an optimal scheduling scheme, so that time consumption of a large number of scheme iteration simulation calculation can be avoided, and real-time optimal scheduling of flood control and drainage of the gate pump group is realized.
Description
Technical Field
The invention relates to the field of emergency disaster prevention, in particular to a method, a system and a storage medium for real-time optimal scheduling of floodgate pump group flood control and drainage.
Background
In the economically developed estuary areas such as Zhujiang delta, changjiang delta and the like in China, the land is flat, the joint is numerous, the river networks are staggered, and the sluice pumps are densely distributed. Gate pump scheduling is one of the important works for guaranteeing flood safety in river network areas. Because the river network area is influenced by the dual dynamic of the radial tide, the water flow condition is complex, and the number of the sluice pumps is numerous, the sluice pump flood control and drainage scheduling of the river network area is determined to be a complex optimization decision problem with multiple dimensions, multiple decision variables and multiple constraints. At present, most river network gate pump group scheduling is still in unordered scheduling or empirical scheduling under the restriction of technical conditions, and the mode is simple and time-saving, but lacks scientificity, so that a global optimal scheduling scheme is difficult to obtain through joint cooperative scheduling among gate pumps; the scheduling model based on hydrodynamic force calculation can solve the problem of optimal scheduling of the gate pump group, but a large number of schemes are needed for iterative simulation calculation, so that the timeliness is poor.
Flood disasters have the characteristics of strong burst and quick disaster, and once the disasters occur, life and property safety of residents can be seriously endangered, and normal and safe operation of cities is influenced. Therefore, in practical application of flood control and drainage scheduling of the gate pump group, the formulation of the scheduling scheme needs to ensure scientificity and timeliness. A river network area gate pump group flood control and drainage real-time optimal scheduling method which combines scientificity and timeliness is sought, and the method has important significance for flood security assurance of a river network area.
Disclosure of Invention
In view of the above problems, the present invention aims to provide a method, a system and a storage medium for real-time optimized scheduling of floodgate pump group flood control and drainage.
The invention provides a floodgate pump group flood control and drainage real-time optimal scheduling method, which comprises the following steps:
s1: taking a region to be scheduled as an object, constructing a hydrodynamic model considering brake pump scheduling;
s2: formulating an evaluation target and a comprehensive evaluation method for controlling a floodgate pump group flood control and drainage scheduling scheme of a region to be scheduled;
s3: selecting boundary conditions and initial conditions of different combinations, generating gate pump group optimization scheduling schemes under different typical situations through trial calculation and evaluation of different gate pump group scheduling schemes based on an evaluation target and a comprehensive evaluation method of gate pump group flood control and drainage scheduling scheme regulation established by S1 and S2, and summarizing to form a scheduling scheme library;
s4: developing decision variable sensitivity analysis on decision variables of the scheduling scheme by adopting a Morris method; screening key decision variables influencing the dispatching decisions of the gate pump group;
s5: grouping the sluice pumps in the area to be scheduled by taking whether the hydraulic connection exists in the river channel as a judging standard;
s6: taking a gate pump group with hydraulic connection as an object, constructing a gate pump scheduling model based on LSTM, taking key decision variable values corresponding to each scheme in a scheduling scheme library, hydrologic boundary conditions and corresponding optimized scheduling schemes as inputs, and carrying out training and testing of the LSTM model;
s7: and generating an optimized dispatching scheme for flood control and drainage of the sluice pump group of the area to be dispatched by using the tested parameters, and then dispatching the sluice pump group of the area to be dispatched in real time by utilizing the optimized dispatching scheme for flood control and drainage of the sluice pump group.
Preferably, the hydrodynamic model considering the scheduling of the brake pump adopts a san veland equation set as a control equation, and the control equation is as follows:
wherein: x is mileage; t is time; z is the water level; b is the water surface width of the water cross section; q is flow; q is lateral single-width flow, (positive value represents inflow and negative value represents outflow), A is water cross-sectional area; g is gravity acceleration; u is the average flow velocity of the section; beta is a correction coefficient; r is the hydraulic radius; c is the thank you coefficient, c=r 1/6 And n, n is the coefficient of Manning roughness.
Preferably, in the hydrodynamic model taking into account the sluice pump schedule, the flux of the sluice section is determined by a sluice overflow equation;
i.e. when the gate is closed, the passing flow q=0;
under the condition of opening the gate, calculating the flow of the gate according to a wide top weir formula:
free outflow:
submerged outflow:
wherein: m is the free flow coefficient;to drown out the flow coefficient; z is Z 0 Is the gate bottom elevation; z is Z u The water level is upstream of the gate; z is Z d Is the downstream water level of the sluice; h 0 The water depth is the water depth at the upstream of the gate; h s Is the water depth downstream of the sluice.
Preferably, the evaluating the target in S2 includes:
(1) a flood control and drainage safety target scored according to the following formula:
wherein S is the ratio of the highest water level of a flood peak of a flood prevention key section to the designed water level; NZ (NZ) Reaching the standard The number of sections of which the highest flood peak water level is lower than the designed water level is calculated; NZ (NZ) Total (S) Is the total number of sections;
(2) landscape water level targets scored according to the following formula:
wherein V is the ratio of the water level of each river representative section not lower than the landscape water level; NJ (NJ) Reaching the standard The number of sections of which the pre-drainage lowest water level is not lower than the landscape water level; NZ (NZ) Total (S) Is the total number of sections;
(3) an operability goal scored according to the following formula:
wherein O is the ratio of the number of water gates for scheduling operation to the total number; NS (NS) Operation of The number of floodgates for the scheduling operation; NS (NS) Total (S) The total number of the sluice gates;
(4) a low carbon energy saving goal scored according to the following formula:
wherein, C is the ratio of the sum of the actual drainage flows of all pump stations to the sum of the design drainage flows; QR (quick response) i The actual drainage flow of the ith pump station; QDs i The drainage flow designed for the ith pump station; p is the total number of pump stations.
Preferably, the comprehensive evaluation method of S2:
the gate pump scheduling scheme is comprehensively evaluated by adopting a gray multi-objective optimization algorithm, wherein the gray multi-objective optimization algorithm consists of an upper limit effect measure and a lower limit effect measure:
the upper limit effect measure formula is:
the lower limit effect measure formula is:
the comprehensive effect measure formula is as follows:
in the method, in the process of the invention,scoring a target k for a scheduling scheme i; />Is->Is a measure of the effect of (1); θ k As the weight of the target k,F i and (5) measuring the comprehensive effect corresponding to the scheme i.
Preferably, the Morris method of S4 has the formula:
wherein S is i (X, deltax) is a sensitivity index of the parameter i; y (x) is the model output result; x= (X 1 ，x 2 ，…，x D ) Is a D-dimensional vector of parameters; Δx is the amount of change in x.
Preferably, S6 comprises the steps of:
s6.1: carrying out data normalization processing on the key decision variables obtained by the S4 screening, integrating the key decision variables into an input array format of an LSTM model, and dividing the data into training set data and test set data according to requirements;
the data normalization method comprises the following steps:
wherein: y is a normalized data value, x is an original data value, min is a minimum value in the sequence, and max is a maximum value in the sequence;
s6.2: creating an LSTM model, and setting key parameters and training parameters of the LSTM model;
s6.3: putting the training set data into an LSTM model for training, if the LSTM model cannot be converged, returning to the step S6.2, and retraining the LSTM model after adjusting key parameters and training parameters of the LSTM model until the loss function reaches the expected value of the loss function;
s6.4: performing model test by adopting LSTM model parameters and test group data obtained through training, and judging a model prediction effect by a mean square error Method (MSE);
if the test does not pass, returning to the step S6.2, and retraining and testing the LSTM model after adjusting the key parameters and the training parameters of the LSTM model until the test passes.
Preferably, the LSTM model is composed of a forgetting door, a memory door, a cell state and an output door, and the main working principles of each part are as follows:
forgetting door f t ＝σ(W f [h t-1 ，x t ]+b f )
Memory gate i t ＝σ(W i [h t-1 ，x t ]+b i )
Cell state c t ＝f t c t-1 +i t tan(W c [h t-1 ，x t ]+b c )
Output door o t ＝σ(W o [h t-1 ，x t ]+b o )
Wherein σ is an activation function sigmoid (); tan is the activation function tanh (); w (W) f A weight matrix for forgetting gates; h is a t-1 The hidden state is the time t-1; x is x t The input is the unit input at the time t; b f A bias vector that is a forget gate; w (W) i A weight matrix for the memory gate; b i A bias vector for the memory gate; w (W) c A weight matrix for the cell state; b c Is a bias vector for the cell state; w (W) o Outputting a weight matrix of the gate; b o Is the bias vector of the output gate.
Preferably, the mean square error Method (MSE) is calculated as follows:
wherein MSE is mean square error, T is test sequence length, i is current sequence number, f (x i ) For LSTM model predictive value, y i To optimize the scheduling scheme values.
The invention provides a floodgate pump group flood control and drainage real-time optimal scheduling system, which is characterized by comprising a memory and a processor, wherein the memory comprises a floodgate pump group flood control and drainage real-time optimal scheduling method program, and the floodgate pump group flood control and drainage real-time optimal scheduling method program is executed by the processor and comprises the following steps:
s1: taking a region to be scheduled as an object, constructing a hydrodynamic model considering brake pump scheduling;
s2: formulating an evaluation target and a comprehensive evaluation method for controlling a floodgate pump group flood control and drainage scheduling scheme of a region to be scheduled;
s3: selecting boundary conditions and initial conditions of different combinations, generating gate pump group optimization scheduling schemes under different typical situations through trial calculation and evaluation of different gate pump group scheduling schemes based on an evaluation target and a comprehensive evaluation method of gate pump group flood control and drainage scheduling scheme regulation established by S1 and S2, and summarizing to form a scheduling scheme library;
s4: developing decision variable sensitivity analysis on decision variables of the scheduling scheme by adopting a Morris method; screening key decision variables influencing the dispatching decisions of the gate pump group;
s5: grouping the sluice pumps in the area to be scheduled by taking whether the hydraulic connection exists in the river channel as a judging standard;
s6: taking a gate pump group with hydraulic connection as an object, constructing a gate pump scheduling model based on LSTM, taking key decision variable values corresponding to each scheme in a scheduling scheme library, hydrologic boundary conditions and corresponding optimized scheduling schemes as inputs, and carrying out training and testing of the LSTM model;
s7: and generating an optimized dispatching scheme for flood control and drainage of the sluice pump group of the area to be dispatched by using the tested parameters, and then dispatching the sluice pump group of the area to be dispatched in real time by utilizing the optimized dispatching scheme for flood control and drainage of the sluice pump group.
The third aspect of the present invention provides a computer readable storage medium, where the computer readable storage medium includes a program of a real-time optimized dispatching method for flood control and drainage of a gate pump group, and when the program of the real-time optimized dispatching method for flood control and drainage of a gate pump group is executed by a processor, the steps of the method for real-time optimized dispatching for flood control and drainage of a gate pump group are implemented.
Compared with the prior art, the technical scheme of the invention has the beneficial effects that: the invention takes a river network of a research area as an object to construct a one-dimensional hydrodynamic model which can consider the dispatching function of a brake pump. And designating a gate pump scheduling scheme comprehensive evaluation scheme. And then, on the basis, boundary conditions and initial conditions of different combinations are selected, gate pump group optimization scheduling schemes under different typical situations are formulated, and a scheduling scheme library is formed by summarizing. And then, carrying out decision variable sensitivity analysis on the decision variable of the scheduling scheme, and carrying out decision variable dimension reduction. And grouping the brake pumps in the to-be-scheduled area, constructing a brake pump scheduling model based on the LSTM, and optimizing a scheduling scheme by using the brake pump scheduling model of the LSTM. The LSTM model is a special RNN (Recurrent Neural Network) model, and the special structural design of the LSTM model enables the LSTM model to consider the influence of the historical time information on the current scheduling scheme. The LSTM model is trained by adopting an optimized scheduling scheme generated by a scheduling model based on hydrodynamic force calculation, so that the LSTM model can learn to obtain the relation between the optimized scheduling scheme and each decision variable. The trained LSTM model is used for generating an optimal scheduling scheme, so that time consumption of a large number of scheme iteration simulation calculation can be avoided, and real-time optimal scheduling of flood control and drainage of the gate pump group is realized.
Drawings
Fig. 1 is a flowchart of a method for real-time optimized scheduling of flood control and drainage of floodgate pump groups according to embodiment 1.
FIG. 2 is a flow chart of constructing an LSTM based brake pump scheduling model.
FIG. 3 is a graph of results of a partial scheme LSTM model test of a one-twenty Yongxi gate.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description. It should be noted that, in the case of no conflict, the embodiments of the present application and the features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those described herein, and therefore the scope of the present invention is not limited to the specific embodiments disclosed below.
Example 1
As shown in fig. 1, the embodiment discloses a real-time optimization scheduling method for flood control and drainage of a gate pump group, which comprises the following steps:
s1: taking a region to be scheduled as an object, constructing a hydrodynamic model considering brake pump scheduling;
s2: formulating an evaluation target and a comprehensive evaluation method for controlling a floodgate pump group flood control and drainage scheduling scheme of a region to be scheduled;
s3: selecting boundary conditions and initial conditions of different combinations, generating gate pump group optimization scheduling schemes under different typical situations through trial calculation and evaluation of different gate pump group scheduling schemes based on an evaluation target and a comprehensive evaluation method of gate pump group flood control and drainage scheduling scheme regulation established by S1 and S2, and summarizing to form a scheduling scheme library;
s4: developing decision variable sensitivity analysis on decision variables of the scheduling scheme by adopting a Morris method; screening key decision variables influencing the dispatching decisions of the gate pump group;
s5: grouping the sluice pumps in the area to be scheduled by taking whether the hydraulic connection exists in the river channel as a judging standard;
s6: taking a gate pump group with hydraulic connection as an object, constructing a gate pump scheduling model based on LSTM, taking key decision variable values corresponding to each scheme in a scheduling scheme library, hydrologic boundary conditions and corresponding optimized scheduling schemes as inputs, and carrying out training and testing of the LSTM model;
s7: and generating an optimized dispatching scheme for flood control and drainage of the sluice pump group of the area to be dispatched by using the tested parameters, and then dispatching the sluice pump group of the area to be dispatched in real time by utilizing the optimized dispatching scheme for flood control and drainage of the sluice pump group.
It should be noted that, in order to solve the problem that the scientificity and timeliness of the traditional gate pump group scheduling method are difficult to be compatible, the application provides a real-time optimization scheduling method for flood control and drainage of a gate pump group. The method mainly takes a river network of a research area as an object, and builds a one-dimensional hydrodynamic model which can consider the dispatching effect of the brake pump. And designating a gate pump scheduling scheme comprehensive evaluation scheme. And then, on the basis, boundary conditions and initial conditions of different combinations are selected, gate pump group optimization scheduling schemes under different typical situations are formulated, and a scheduling scheme library is formed by summarizing. And then, carrying out decision variable sensitivity analysis on the decision variable of the scheduling scheme, and carrying out decision variable dimension reduction. And grouping the brake pumps in the to-be-scheduled area, constructing a brake pump scheduling model based on the LSTM, and optimizing a scheduling scheme by using the brake pump scheduling model of the LSTM.
The LSTM model in the present application is a special RNN (Recurrent Neural Network) model, and its special structural design makes it possible to consider the influence of the historical time information on the current scheduling scheme. The LSTM model is trained by adopting an optimized scheduling scheme generated by a scheduling model based on hydrodynamic force calculation, so that the LSTM model can learn to obtain the relation between the optimized scheduling scheme and each decision variable. The trained LSTM model is used for generating an optimal scheduling scheme, so that time consumption of a large number of scheme iteration simulation calculation can be avoided, and real-time optimal scheduling of flood control and drainage of the gate pump group is realized.
According to the embodiment of the invention, the hydrodynamic model considering the dispatching of the brake pump adopts the san View equation set as a control equation, and the control equation is as follows:
wherein: x is mileage; t is time; z is the water level; b is the water surface width of the water cross section; q is flow; q is lateral single-width flow, (positive value represents inflow and negative value represents outflow), A is water cross-sectional area; g is gravity acceleration; u is the average flow velocity of the section; beta is a correction coefficient; r is the hydraulic radius; c is the thank you coefficient, c=r 1/6 And n, n is the coefficient of Manning roughness.
According to the embodiment of the invention, in a hydrodynamic model considering sluice pump scheduling, the flux of a sluice section is determined by a sluice overflow formula;
i.e. when the gate is closed, the passing flow q=0;
under the condition of opening the gate, calculating the flow of the gate according to a wide top weir formula:
free outflow:
submerged outflow:
wherein: q isThe flow rate of the passing gate; m is the free flow coefficient;to drown out the flow coefficient; b is the total width of the gate opening; z is Z 0 Is the gate bottom elevation; z is Z u The water level is upstream of the gate; z is Z d Is the downstream water level of the sluice; h 0 The water depth is the water depth at the upstream of the gate; h s Is the water depth downstream of the sluice.
Note that, in consideration of the hydrodynamic model of the sluice pump scheduling, the flow Q in the above formula substantially corresponds to the sluice flow, and the water surface width B of the water section substantially corresponds to the total width of the sluice opening.
According to the method, the hydrodynamic model for scheduling the sluice pump is considered, parameters such as the opening and closing state of the sluice, the number of holes, the opening and the like are determined according to the sluice scheduling rule, and then the sluice flow is calculated, so that the sluice engineering scheduling process is reasonably generalized.
For the pump station, the pump station is generalized into side inflow/outflow items in the river network model, so that the influence of pump station water pumping and draining on river water quantity is reflected. The model generalizes pump station dispatching rules through parameters such as a starting water level, a stopping water level, pump station pumping flow and the like.
According to an embodiment of the present invention, the step S2 of evaluating the target includes:
(1) a flood control and drainage safety target scored according to the following formula:
wherein S is the ratio of the highest water level of a flood peak of a flood prevention key section to the designed water level; NZ (NZ) Reaching the standard The number of sections of which the highest flood peak water level is lower than the designed water level is calculated; NZ (NZ) Total (S) Is the total number of sections.
The flood control and drainage safety aims at controlling the water level of the critical flood section and the highest ratio of the highest water level of the critical flood section to the highest water level of the flood peak to the designed water level.
(2) Landscape water level targets scored according to the following formula:
wherein V is the ratio of the water level of each river representative section not lower than the landscape water level; NJ (NJ) Reaching the standard The number of sections of which the pre-drainage lowest water level is not lower than the landscape water level; NZ (NZ) Total (S) Is the total number of sections.
It should be noted that, because the landscape water level is required to be the highest in the river pre-drainage process, the lowest water level of each river representative section is not lower than the highest ratio of the landscape water level. Typically this index is considered when the basin encounters standard internal rainfall and is not considered when an oversubstance rainfall occurs.
(3) An operability goal scored according to the following formula:
wherein O is the ratio of the number of water gates for scheduling operation to the total number; NS (NS) Operation of The number of floodgates for the scheduling operation; NS (NS) Total (S) Is the total number of sluice gates.
Since the number of gates in the river network area is large, the ratio of the number of gates to the total number of the scheduling operations is the lowest in order to achieve the same object from the viewpoint of the operability of the scheme.
(4) A low carbon energy saving goal scored according to the following formula:
wherein, C is the ratio of the sum of the actual drainage flows of all pump stations to the sum of the design drainage flows; QR (quick response) i The actual drainage flow of the ith pump station; QDs i The drainage flow designed for the ith pump station; p is the total number of pump stations.
It should be noted that, because the energy consumption of the drainage pump station is larger, from the viewpoint of low carbon and energy saving, the lowest ratio of the total actual drainage flow of all pump stations to the total design drainage flow is taken as the target under the condition of realizing the same target.
According to the embodiment of the invention, the comprehensive evaluation method in S2 is as follows:
the gate pump scheduling scheme is comprehensively evaluated by adopting a gray multi-objective optimization algorithm, wherein the gray multi-objective optimization algorithm consists of an upper limit effect measure and a lower limit effect measure:
the upper limit effect measure formula is:
the lower limit effect measure formula is:
the comprehensive effect measure formula is as follows:
in the method, in the process of the invention,scoring a target k for a scheduling scheme i; />Is->Is a measure of the effect of (1); θ k As the weight of the target k,F i and (5) measuring the comprehensive effect corresponding to the scheme i.
In addition, it should be noted that, in the present application, step S3 selects a plurality of sets of typical boundary conditions and initial conditions of different combinations, including a flow (rainfall) process of the research area, an out river (sea) water (tide) process, an initial state of a river channel, a sluice, a pump station, and the like.
According to the embodiment of the present invention, the formula of the Morris method described in S4 is:
wherein S is i (X, deltax) is a sensitivity index of the parameter i; y (x) is the model output result; x= (X 1 ，x 2 ，…，x D ) Is a D-dimensional vector of parameters; Δx is the amount of change in x.
It should be noted that, because of numerous factors affecting the flood control and drainage scheduling scheme of the sluice pump in the river network area, besides the critical influence of boundary conditions such as flow (or rainfall) process, tide level process, etc. on the formulation of the sluice pump group scheduling scheme, various factors such as the early state of the river basin (such as soil saturation degree, confluence yielding condition), the early state of the river channel (such as water level, flow rate, flow velocity), the initial state of the sluice pump station (such as the opening and closing state of the sluice pump station, the number of sluice opening holes, the gate opening degree, and the pump station flow rate) all affect the formulation of the flood control and drainage scheduling scheme of the sluice pump. The factors are mutually influenced to enable the scheduling decision dimension to be exponentially increased. If the giant dimension decision variable is used as the input of the deep learning model, the model can not be converged, and the function of the model can not be exerted. Therefore, it is necessary to analyze the contribution rate of each decision variable to the influence of the gate pump group scheduling scheme by performing sensitivity analysis on the decision variables, and screen out the key decision variables influencing the gate pump group scheduling decision. And carrying out decision variable sensitivity analysis by adopting a Morris method.
According to an embodiment of the present invention, as shown in fig. 2, S6 includes the steps of:
s6.1: carrying out data normalization processing on the key decision variables obtained by the S4 screening, integrating the key decision variables into an input array format of an LSTM model, and dividing the data into training set data and test set data according to requirements;
the data normalization method comprises the following steps:
wherein: y is the normalized data value, x is the original data value, min is the minimum value in the sequence, and max is the maximum value in the sequence.
Before constructing the brake pump scheduling model based on the LSTM, preprocessing the key variable value hydrologic boundary condition obtained by S4 analysis and the corresponding optimized scheduling scheme, namely normalizing the data, and integrating the data after normalizing the data into the input array format of the LSTM model.
S6.2: creating an LSTM model, and setting key parameters and training parameters of the LSTM model.
It should be noted that, the key parameters of the LSTM model include an input feature dimension, a feature dimension of the hidden layer, an output feature dimension, a number of layers of the hidden layer, and the training parameters include a learning rate, a maximum iteration number, an expected value of a loss function, and the like.
S6.3: putting the training set data into an LSTM model for training, if the LSTM model cannot be converged, returning to the step S6.2, and retraining the LSTM model after adjusting key parameters and training parameters of the LSTM model until the loss function reaches the expected value of the loss function;
s6.4: performing model test by adopting LSTM model parameters and test group data obtained through training, and judging a model prediction effect by a mean square error Method (MSE);
if the test does not pass, returning to the step S6.2, and retraining and testing the LSTM model after adjusting the key parameters and the training parameters of the LSTM model until the test passes.
According to the embodiment of the invention, the LSTM model consists of a forgetting gate, a memory gate, a cell state and an output gate, and the main working principles of the parts are as follows:
forgetting door f t ＝σ(W f [h t-1 ，x t ]+b f )
Memory gate i t ＝σ(W i [h t-1 ，x t ]+b i )
Cell state c t ＝f t c t-1 +i t tan(W c [h t-1 ，x t ]+b c )
Output door o t ＝σ(W o [h t-1 ，x t ]+b o )
Wherein σ is an activation function sigmoid (); tan is an activation functiontanh()；W f A weight matrix for forgetting gates; h is a t-1 The hidden state is the time t-1; x is x t The input is the unit input at the time t; b f A bias vector that is a forget gate; w (W) i A weight matrix for the memory gate; b i A bias vector for the memory gate; w (W) c A weight matrix for the cell state; b c Is a bias vector for the cell state; w (W) o Outputting a weight matrix of the gate; b o Is the bias vector of the output gate.
According to an embodiment of the present invention, the mean square error Method (MSE) is calculated as follows:
wherein MSE is mean square error, T is test sequence length, i is current sequence number, f (x i ) For LSTM model predictive value, y i To optimize the scheduling scheme values.
As a specific example, to verify the feasibility and effectiveness of the present application, the present example is further described with respect to a specific implementation of the generation of a sand gate pump scheduling scheme for ten thousand hectares in southern sand region, guangzhou. The Sha-Lian-Jie is located in the southwest of New Nansha district in Guangzhou, the north side is lower horizontal drainage, the east side is Dragon cave south water channel, the west side is Hong Ji drainage channel, and the south side faces the sea. The total length of 33 existing rivers in the surrounding area is about 148km, and one to eighteen surges in the sheet area are flood drainage bone dry rivers. A total of 48 water gates are distributed in the sheet area, and are all tidal gates of the outer river, and the total clear width is 637.2m; the total design flow rate of the outer pumping station is 56.76m, and the outer pumping station is 1 seat 3 /s。
The floodgate pump group flood control and drainage real-time optimization scheduling method of the embodiment is utilized to schedule the floodgate pump group flood control and drainage of the sand surrounding the ten thousand hectares.
S1: a one-dimensional river network model which considers 33 inland river surges, 48 sluice gates and 1 pump station around the sand of ten thousand hectares is constructed, and 523 river sections are arranged. Considering that the river around the inner river is dense, the water collecting area is small and the converging speed is high. The runoff is generalized in a runoff coefficient mode, and the converging time is set to be 20 minutes.
S2: setting a ten thousand-hectare sand surrounding gate pump group flood control and drainage scheduling scheme evaluation target. The flood control and drainage safety target is characterized in that 21 key sections are set by taking sections of important river crossing points and dense areas of urban areas as key sections; the landscape water level target takes the current state pre-dewatering level of ten thousand hectares of sand as the landscape water level target, wherein the pre-dewatering level is 4.5m (the height of the urban construction in Guangzhou). Operability targets, the lowest ratio of the number of water gates to the total number for scheduling operations. And the low-carbon energy-saving target is to take the minimum starting power of the two-welling-western pump station as a target. A gray multi-objective optimization algorithm is adopted to comprehensively evaluate the advantages and disadvantages of the gate pump scheduling scheme.
S3: the boundary conditions are selected from 5-100 years of 5-reproduction-period design rainfall and 5 ten thousand-hectare sand history typical rain, 5-100 years of 5-reproduction-period design tide level and 4 (big, medium, small tide and typical storm tide) typical tide level processes of ten thousand-hectare sand stations; the initial conditions are set, the initial water level of the inland river in the surrounding area adopts a unified value, and the value is taken from 4.5m to 6.5m at intervals of 0.5m, and 5 initial water level conditions are adopted. The inner water gates are all single Kong Shuizha, and for simplicity, it is assumed that the initial states of all water gates are kept consistent, and 2 states of full open and full closed are adopted. The two-gushing western pump station is started and closed and is associated with the two-gushing western gate, and when the two-gushing western gate is closed, the two-gushing western pump station is fully opened. Boundary conditions and initial conditions combined to form a total of 5000 groups of typical calculation schemes of 5 x 4 x 5 x 2 x 1, and calculating by a gate pump group scheduling model to form a sand surrounding gate pump joint scheduling scheme library for ten thousand hectares.
S4: and carrying out decision variable sensitivity analysis. In this embodiment, the production convergence process simulation is integrated in the one-dimensional model, and the drainage basin early state is considered in the scheduling model. The initial state of the river channel considers the initial water level of the river channel as a decision variable, the initial state of the sluice pump station considers the opening and closing state of the sluice as a decision variable, and the initial state of the two-gushing pump station is not considered as the decision variable due to the association with the opening and closing state of the sluice. And a Morris method is adopted to carry out decision variable sensitivity analysis, and a scheduling scheme is insensitive to the initial opening and closing state of the sluice through analysis. Therefore, only one key decision variable affecting the scheduling decision of the gate pump group is set, namely the initial water level of the river channel.
S5: and (3) grouping sluice pump stations around the sand of ten hectares by taking whether the water power connection exists in the river channel as a judgment standard. A total of 9 packets were divided and the packet results are shown in the following table.
S6: and taking each group as an object, constructing a brake pump scheduling model based on the LSTM, and carrying out training and testing of the LSTM model. For convenience of explanation, the LSTM model construction process is further described with reference to group 9 (twenty-one Yongdong gate, twenty-one Yongxi gate).
S6.1, taking the initial water level, rainfall process, tide level process and the corresponding optimized scheduling scheme corresponding to each scheme in the scheduling scheme library as input, carrying out data normalization processing, and integrating the data normalization processing into an input array format of the LSTM model. Of 5000 groups of typical calculation schemes, 80% (4000 groups) were used as training data and 20% (1000 groups) were used as test data.
S6.2, creating an LSTM model, setting the input characteristic dimension to 4 (rainfall, tide level, initial water level before the twenty-first Yongdong gate and initial water level before the twenty-first Yongxi gate), setting the characteristic dimension of the hidden layer to 100, setting the output characteristic dimension to 2 (the twenty-first Yongdong gate and twenty-first Yongxi gate scheduling schemes), setting the hidden layer number to 2, setting the learning rate to 0.01, setting the maximum iteration number to 10000, and setting the expected value of the loss function to < 1×10 -4 。
S6.3, after setting initial parameters of the LSTM model, putting test group data into the LSTM model for training, if the model cannot be converged, returning to S6.2 again, and retraining after parameter adjustment until the loss function reaches an expected value of the loss function.
And S6.4, carrying out model test by using the model parameters obtained by training and the test group data, and judging the model prediction effect by using a mean square error Method (MSE). If the test does not pass, returning to S6.2 again, and retraining and testing after parameter adjustment until the test passes. Through adjustment, the final key parameter value of the LSTM model is that the characteristic dimension of the hidden layer is adjusted to be 200, the number of the hidden layer is maintained to be 2, and the learning rate is improvedAdjusting to 0.001, maintaining the maximum iteration number to 10000, and adjusting the expected value of the loss function to < 1×10 -5 。
From the test effect, under the simplified condition of fully opening and fully closing only the gate, the mean square error of the generated scheduling scheme and the result of the scheduling model calculated based on hydrodynamic force is smaller than 0.05 (1 is fully opening and 0 is closing), and the test result of the twenty-first Yongxi gate part scheme is shown in figure 3. The generated scheduling scheme of the brake pump scheduling model based on the LSTM constructed at the time is basically consistent with the result of the calculation scheduling model based on the hydrodynamic force from the test effect, and the method can be used for generating the brake pump optimized scheduling scheme in real time.
S7: and (3) using the parameters passing the test to generate the optimized scheduling scheme for flood control and drainage of the gate pump group. And then, utilizing an optimized dispatching scheme for flood control and drainage of the sluice pump group, and dispatching the sluice pump group of the area to be dispatched in real time.
Example 2
The embodiment discloses a floodgate pump group flood control and drainage real-time optimal scheduling system, which comprises a memory and a processor, wherein the memory comprises a floodgate pump group flood control and drainage real-time optimal scheduling method program, and the floodgate pump group flood control and drainage real-time optimal scheduling method program is executed by the processor to realize the following steps:
s1: taking a region to be scheduled as an object, constructing a hydrodynamic model considering brake pump scheduling;
s2: formulating an evaluation target and a comprehensive evaluation method for controlling a floodgate pump group flood control and drainage scheduling scheme of a region to be scheduled;
s3: selecting boundary conditions and initial conditions of different combinations, generating gate pump group optimization scheduling schemes under different typical situations through trial calculation and evaluation of different gate pump group scheduling schemes based on an evaluation target and a comprehensive evaluation method of gate pump group flood control and drainage scheduling scheme regulation established by S1 and S2, and summarizing to form a scheduling scheme library;
s4: developing decision variable sensitivity analysis on decision variables of the scheduling scheme by adopting a Morris method; screening key decision variables influencing the dispatching decisions of the gate pump group;
s5: grouping the sluice pumps in the area to be scheduled by taking whether the hydraulic connection exists in the river channel as a judging standard;
s6: taking a gate pump group with hydraulic connection as an object, constructing a gate pump scheduling model based on LSTM, taking key decision variable values corresponding to each scheme in a scheduling scheme library, hydrologic boundary conditions and corresponding optimized scheduling schemes as inputs, and carrying out training and testing of the LSTM model;
s7: and generating an optimized dispatching scheme for flood control and drainage of the sluice pump group of the area to be dispatched by using the tested parameters, and then dispatching the sluice pump group of the area to be dispatched in real time by utilizing the optimized dispatching scheme for flood control and drainage of the sluice pump group.
Example 3
The embodiment discloses a computer readable storage medium, which comprises a gate pump group flood control and drainage real-time optimal scheduling method program, wherein when the gate pump group flood control and drainage real-time optimal scheduling method program is executed by a processor, the steps of the gate pump group flood control and drainage real-time optimal scheduling method are realized.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.
The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present invention may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, where the program, when executed, performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk or an optical disk, or the like, which can store program codes.
Alternatively, the above-described integrated units of the present invention may be stored in a computer-readable storage medium if implemented in the form of software functional modules and sold or used as separate products. Based on such understanding, the technical solutions of the embodiments of the present invention may be embodied in essence or a part contributing to the prior art in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, ROM, RAM, magnetic or optical disk, or other medium capable of storing program code.
Claims (10)
1. A floodgate pump group flood control and drainage real-time optimal scheduling method is characterized by comprising the following steps:
s1: taking a region to be scheduled as an object, constructing a hydrodynamic model considering brake pump scheduling;
s2: formulating an evaluation target and a comprehensive evaluation method for controlling a floodgate pump group flood control and drainage scheduling scheme of a region to be scheduled;
s3: selecting boundary conditions and initial conditions of different combinations, generating gate pump group optimization scheduling schemes under different typical situations through trial calculation and evaluation of different gate pump group scheduling schemes based on an evaluation target and a comprehensive evaluation method of gate pump group flood control and drainage scheduling scheme regulation established by S1 and S2, and summarizing to form a scheduling scheme library;
s4: developing decision variable sensitivity analysis on decision variables of the scheduling scheme by adopting a Morris method; screening key decision variables influencing the dispatching decisions of the gate pump group;
s5: grouping the sluice pumps in the area to be scheduled by taking whether the hydraulic connection exists in the river channel as a judging standard;
s6: taking a gate pump group with hydraulic connection as an object, constructing a gate pump scheduling model based on LSTM, taking key decision variable values corresponding to each scheme in a scheduling scheme library, hydrologic boundary conditions and corresponding optimized scheduling schemes as inputs, and carrying out training and testing of the LSTM model;
s7: and generating an optimized dispatching scheme for flood control and drainage of the sluice pump group of the area to be dispatched by using the tested parameters, and then dispatching the sluice pump group of the area to be dispatched in real time by utilizing the optimized dispatching scheme for flood control and drainage of the sluice pump group.
2. The method for real-time optimal scheduling of floodgate pump group flood control and drainage according to claim 1, wherein the hydrodynamic model taking the floodgate pump scheduling into consideration adopts a san veland equation set as a control equation, and the control equation is as follows:
wherein: x is mileage; t is time; z is the water level; b is the water surface width of the water cross section; q is flow; q is lateral single-width flow, A is water cross-section area; g is gravity acceleration; u is the average flow velocity of the section; beta is a correction coefficient; r is the hydraulic radius; c is the thank you coefficient, c=r 1/6 N, n is a Manning roughness coefficient;
in a hydrodynamic model considering sluice pump scheduling, the flux of sluice sections is determined by a sluice overflow formula;
i.e. when the gate is closed, the passing flow q=0;
under the condition of opening the gate, calculating the flow of the gate according to a wide top weir formula:
free outflow:
submerged outflow:
wherein: m is the free flow coefficient;to drown out the flow coefficient; z is Z 0 Is the gate bottom elevation; z is Z u The water level is upstream of the gate; z is Z d Is the downstream water level of the sluice; h 0 The water depth is the water depth at the upstream of the gate; h s Is the water depth downstream of the sluice.
3. The method for real-time optimal scheduling of floodgate pump group flood control and drainage according to claim 1 or 2, wherein the step S2 of evaluating the target comprises:
(1) a flood control and drainage safety target scored according to the following formula:
wherein S is that the highest water level of flood peak of critical flood prevention section is lower than design waterBit ratio; NZ (NZ) Reaching the standard The number of sections of which the highest flood peak water level is lower than the designed water level is calculated; NZ (NZ) Total (S) Is the total number of sections;
(2) landscape water level targets scored according to the following formula:
wherein V is the ratio of the water level of each river representative section not lower than the landscape water level; NJ (NJ) Reaching the standard The number of sections of which the pre-drainage lowest water level is not lower than the landscape water level; NZ (NZ) Total (S) Is the total number of sections;
(3) an operability goal scored according to the following formula:
wherein O is the ratio of the number of water gates for scheduling operation to the total number; NS (NS) Operation of The number of floodgates for the scheduling operation; NS (NS) Total (S) The total number of the sluice gates;
(4) a low carbon energy saving goal scored according to the following formula:
wherein, C is the ratio of the sum of the actual drainage flows of all pump stations to the sum of the design drainage flows; QR (quick response) i The actual drainage flow of the ith pump station; QDs i The drainage flow designed for the ith pump station; p is the total number of pump stations.
4. The method for real-time optimal scheduling of flood control and drainage of floodgate pump groups according to claim 3, wherein the comprehensive evaluation method of S2 comprises the following steps:
the gate pump scheduling scheme is comprehensively evaluated by adopting a gray multi-objective optimization algorithm, wherein the gray multi-objective optimization algorithm consists of an upper limit effect measure and a lower limit effect measure:
the upper limit effect measure formula is:
the lower limit effect measure formula is:
the comprehensive effect measure formula is as follows:
in the method, in the process of the invention,scoring a target k for a scheduling scheme i; />Is->Is a measure of the effect of (1); θ k As the weight of the target k,F i and (5) measuring the comprehensive effect corresponding to the scheme i.
5. The real-time optimal scheduling method for flood control and drainage of a floodgate pump group according to claim 4, wherein the formula of the Morris method in S4 is as follows:
wherein S is i (X, deltax) is a sensitivity index of the parameter i; y (x) is the model output result; x= (X 1 ，x 2 ，…，x D ]Is a D-dimensional vector of parameters; Δx is the amount of change in x.
6. The method for real-time optimal scheduling of floodgate pump group flood control and drainage according to claim 1 or 5, wherein S6 comprises the following steps:
s6.1: carrying out data normalization processing on the key decision variables obtained by the S4 screening, integrating the key decision variables into an input array format of an LSTM model, and dividing the data into training set data and test set data according to requirements;
s6.2: creating an LSTM model, and setting key parameters and training parameters of the LSTM model;
s6.3: putting the training set data into an LSTM model for training, if the LSTM model cannot be converged, returning to the step S6.2, and retraining the LSTM model after adjusting key parameters and training parameters of the LSTM model until the loss function reaches the expected value of the loss function;
s6.4: performing model test by adopting LSTM model parameters and test group data obtained through training, and judging a model prediction effect by a mean square error method;
if the test does not pass, returning to the step S6.2, and retraining and testing the LSTM model after adjusting the key parameters and the training parameters of the LSTM model until the test passes.
7. The real-time optimal scheduling method for flood control and drainage of a gate pump group according to claim 6, wherein the LSTM model is composed of a forgetting gate, a memory gate, a cell state and an output gate, and the main working principles of the parts are as follows:
forgetting door f t ＝σ(W f [h t-1 ，x t ]+b f )
Memory gate i t ＝σ(W i [h t-1 ，x t ]+b i )
Cell state c t ＝f t c t-1 +i t tan(W c [h t-1 ，x t ]+b c )
Output door o t ＝σ(W o [h t-1 ，x t ]+b o )
Wherein σ is an activation function sigmoid (); tan is the activation function tanh (); w (W) f A weight matrix for forgetting gates; h is a t-1 The hidden state is the time t-1; x is x t The input is the unit input at the time t; b f A bias vector that is a forget gate; w (W) i A weight matrix for the memory gate; b i A bias vector for the memory gate; w (W) c A weight matrix for the cell state; b c Is a bias vector for the cell state; w (W) o Outputting a weight matrix of the gate; b o Is the bias vector of the output gate.
8. The method for real-time optimal scheduling of flood control and drainage of a gate pump group according to claim 7, wherein the mean square error method is calculated as follows:
wherein MSE is mean square error, T is test sequence length, i is current sequence number, f (x i ) For LSTM model predictive value, y i To optimize the scheduling scheme values.
9. The floodgate pump group flood control and drainage real-time optimal scheduling system is characterized by comprising a memory and a processor, wherein the memory comprises a floodgate pump group flood control and drainage real-time optimal scheduling method program, and the floodgate pump group flood control and drainage real-time optimal scheduling method program is executed by the processor to realize the following steps:
s1: taking a region to be scheduled as an object, constructing a hydrodynamic model considering brake pump scheduling;
s2: formulating an evaluation target and a comprehensive evaluation method for controlling a floodgate pump group flood control and drainage scheduling scheme of a region to be scheduled;
s3: selecting boundary conditions and initial conditions of different combinations, generating gate pump group optimization scheduling schemes under different typical situations through trial calculation and evaluation of different gate pump group scheduling schemes based on an evaluation target and a comprehensive evaluation method of gate pump group flood control and drainage scheduling scheme regulation established by S1 and S2, and summarizing to form a scheduling scheme library;
s4: developing decision variable sensitivity analysis on decision variables of the scheduling scheme by adopting a Morris method; screening key decision variables influencing the dispatching decisions of the gate pump group;
s5: grouping the sluice pumps in the area to be scheduled by taking whether the hydraulic connection exists in the river channel as a judging standard;
s6: taking a gate pump group with hydraulic connection as an object, constructing a gate pump scheduling model based on LSTM, taking key decision variable values corresponding to each scheme in a scheduling scheme library, hydrologic boundary conditions and corresponding optimized scheduling schemes as inputs, and carrying out training and testing of the LSTM model;
s7: and generating an optimized dispatching scheme for flood control and drainage of the sluice pump group of the area to be dispatched by using the tested parameters, and then dispatching the sluice pump group of the area to be dispatched in real time by utilizing the optimized dispatching scheme for flood control and drainage of the sluice pump group.
10. A computer readable storage medium, wherein the computer readable storage medium includes a gate pump group flood control and drainage real-time optimal scheduling method program, and when the gate pump group flood control and drainage real-time optimal scheduling method program is executed by a processor, the steps of the gate pump group flood control and drainage real-time optimal scheduling method according to any one of claims 1 to 8 are implemented.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310283301.9A CN116562537B (en) | 2023-03-22 | 2023-03-22 | Floodgate pump group flood control and drainage real-time optimal scheduling method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310283301.9A CN116562537B (en) | 2023-03-22 | 2023-03-22 | Floodgate pump group flood control and drainage real-time optimal scheduling method, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116562537A true CN116562537A (en) | 2023-08-08 |
| CN116562537B CN116562537B (en) | 2023-10-31 |
Family
ID=87498966
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310283301.9A Active CN116562537B (en) | 2023-03-22 | 2023-03-22 | Floodgate pump group flood control and drainage real-time optimal scheduling method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116562537B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117454784A (en) * | 2023-10-07 | 2024-01-26 | 珠江水利委员会珠江水利科学研究院 | Water gate tide-blocking and waterlogging-draining joint scheduling dimension-reducing method, system and storage medium |
| CN117494616A (en) * | 2023-12-28 | 2024-02-02 | 珠江水利委员会珠江水利科学研究院 | Gate pump group scheduling simulation method and system |
| CN118153734A (en) * | 2024-01-19 | 2024-06-07 | 水利部水利水电规划设计总院 | Multi-objective optimization scheduling rule extraction method for water network regulation engineering group |
| CN118313641A (en) * | 2024-06-12 | 2024-07-09 | 珠江水利委员会珠江水利科学研究院 | Gate pump group multi-target cooperative balance scheduling method, system and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016161285A (en) * | 2015-02-26 | 2016-09-05 | 株式会社荏原製作所 | Liquid pump maintenance scheduler |
| CN106202618A (en) * | 2016-06-24 | 2016-12-07 | 珠江水利委员会珠江水利科学研究院 | A kind of Project Scheduling and the method for numerical simulation of tidal river network pollutant defeated shifting PROCESS COUPLING |
| KR101864342B1 (en) * | 2018-02-07 | 2018-06-04 | 서울시립대학교 산학협력단 | Method for Optimal Water Supply Pump Operation Based on Short-term Water Demand Forecasting Considering Disinfection Performance in Clearwell |
| CN109872063A (en) * | 2019-02-11 | 2019-06-11 | 南昌工程学院 | The flood control of the plain city network of waterways, water drainage, running water joint optimal operation method and system |
| JP2020013581A (en) * | 2019-08-01 | 2020-01-23 | 株式会社荏原製作所 | Liquid pump maintenance scheduler |
| CN113763204A (en) * | 2021-08-31 | 2021-12-07 | 中冶华天工程技术有限公司 | Method for evaluating water environment improvement effect of river network water regulation engineering in plain area under multi-objective optimization |
| CN113850692A (en) * | 2021-09-26 | 2021-12-28 | 天津大学 | Urban water supply system gate pump group optimal scheduling method based on deep learning |
| CN114357868A (en) * | 2021-12-22 | 2022-04-15 | 武汉大学 | Multi-target cooperative scheduling method and device for complex flood control system |
| CN114792071A (en) * | 2022-05-18 | 2022-07-26 | 西安理工大学 | Optimal scheduling method for drainage pump station based on machine learning technology |
-
2023
- 2023-03-22 CN CN202310283301.9A patent/CN116562537B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016161285A (en) * | 2015-02-26 | 2016-09-05 | 株式会社荏原製作所 | Liquid pump maintenance scheduler |
| CN106202618A (en) * | 2016-06-24 | 2016-12-07 | 珠江水利委员会珠江水利科学研究院 | A kind of Project Scheduling and the method for numerical simulation of tidal river network pollutant defeated shifting PROCESS COUPLING |
| KR101864342B1 (en) * | 2018-02-07 | 2018-06-04 | 서울시립대학교 산학협력단 | Method for Optimal Water Supply Pump Operation Based on Short-term Water Demand Forecasting Considering Disinfection Performance in Clearwell |
| CN109872063A (en) * | 2019-02-11 | 2019-06-11 | 南昌工程学院 | The flood control of the plain city network of waterways, water drainage, running water joint optimal operation method and system |
| JP2020013581A (en) * | 2019-08-01 | 2020-01-23 | 株式会社荏原製作所 | Liquid pump maintenance scheduler |
| CN113763204A (en) * | 2021-08-31 | 2021-12-07 | 中冶华天工程技术有限公司 | Method for evaluating water environment improvement effect of river network water regulation engineering in plain area under multi-objective optimization |
| CN113850692A (en) * | 2021-09-26 | 2021-12-28 | 天津大学 | Urban water supply system gate pump group optimal scheduling method based on deep learning |
| CN114357868A (en) * | 2021-12-22 | 2022-04-15 | 武汉大学 | Multi-target cooperative scheduling method and device for complex flood control system |
| CN114792071A (en) * | 2022-05-18 | 2022-07-26 | 西安理工大学 | Optimal scheduling method for drainage pump station based on machine learning technology |
Non-Patent Citations (4)
| Title |
|---|
| ZHAO CHENGPING: "Research and Design of Distributed Sluice Group Joint Dispatch System Based on WCF and OPC Technology", 《2010 SECOND WORLD CONGRESS ON SOFTWARE ENGINEERING》, pages 37 - 40 * |
| 胡晓张: "基于水系联通的珠三角典型联围闸泵群调度方案研究", 人民珠江, vol. 41, no. 05, pages 101 - 107 * |
| 闫福恩: "滨海平原河网闸群联合调度模拟", 南昌大学学报(工科版), vol. 42, no. 01, pages 51 - 56 * |
| 魏良良: "基于BP神经网络与改进遗传算法的泵站优化调度", 水电能源科学, vol. 37, no. 05, pages 168 - 171 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117454784A (en) * | 2023-10-07 | 2024-01-26 | 珠江水利委员会珠江水利科学研究院 | Water gate tide-blocking and waterlogging-draining joint scheduling dimension-reducing method, system and storage medium |
| CN117454784B (en) * | 2023-10-07 | 2024-06-11 | 珠江水利委员会珠江水利科学研究院 | Water gate tide-blocking and waterlogging-draining joint scheduling dimension-reducing method, system and storage medium |
| CN117494616A (en) * | 2023-12-28 | 2024-02-02 | 珠江水利委员会珠江水利科学研究院 | Gate pump group scheduling simulation method and system |
| CN117494616B (en) * | 2023-12-28 | 2024-04-26 | 珠江水利委员会珠江水利科学研究院 | Gate pump group scheduling simulation method and system |
| CN118153734A (en) * | 2024-01-19 | 2024-06-07 | 水利部水利水电规划设计总院 | Multi-objective optimization scheduling rule extraction method for water network regulation engineering group |
| CN118313641A (en) * | 2024-06-12 | 2024-07-09 | 珠江水利委员会珠江水利科学研究院 | Gate pump group multi-target cooperative balance scheduling method, system and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116562537B (en) | 2023-10-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116562537B (en) | Floodgate pump group flood control and drainage real-time optimal scheduling method, system and storage medium | |
| Kisi et al. | River suspended sediment estimation by climatic variables implication: Comparative study among soft computing techniques | |
| Jeong et al. | Rainfall‐runoff models using artificial neural networks for ensemble streamflow prediction | |
| Firat et al. | River flow estimation using adaptive neuro fuzzy inference system | |
| Karimi et al. | Neuro-fuzzy and neural network techniques for forecasting sea level in Darwin Harbor, Australia | |
| KR102236678B1 (en) | Method and device for forecasting flood based on data analyzing | |
| CN108109076B (en) | Method for analyzing risk of abandoned water in power generation dispatching of cascade hydropower station group by considering runoff forecasting | |
| CN117236673B (en) | Urban water network multi-scale flood control and drainage combined optimization scheduling method and system | |
| Kabiri-Samani et al. | Application of neural networks and fuzzy logic models to long-shore sediment transport | |
| CN108681844B (en) | Flood resource utilization risk evaluation method for water diversion project | |
| CN114492233B (en) | Watershed water simulation method based on webGIS platform and considering comprehensive utilization requirements | |
| CN118009990B (en) | High-precision real-time forecasting method for tide level based on transducer model | |
| CN118095104B (en) | Machine learning-based flood rapid forecasting method and system | |
| Vafakhah et al. | Application of intelligent technology in rainfall analysis | |
| Park et al. | Application of recurrent neural network for inflow prediction into multi-purpose dam basin | |
| Chen et al. | Evaluation of hybrid soft computing model’s performance in estimating wave height | |
| Xu et al. | Predicting wave forces on coastal bridges using genetic algorithm enhanced ensemble learning framework | |
| Zanganeh et al. | ANFIS and ANN models for the estimation of wind and wave-induced current velocities at Joeutsu-Ogata coast | |
| Zhao et al. | Forewarning model for water pollution risk based on Bayes theory | |
| Mahmoud et al. | Artificial Neural Network Model for Forecasting Haditha Reservoir Inflow in the West of Iraq | |
| CN114519308A (en) | Method for determining river water and underground water interconversion lag response time influenced by river water and sand regulation | |
| CN115659781A (en) | Coastal gate station tide level prediction method and system | |
| Guru et al. | Application of soft computing techniques for river flow prediction in the downstream catchment of Mahanadi River Basin using partial duration series, India | |
| Yan et al. | Data-driven modeling of sluice gate flows using a convolutional neural network | |
| Nguyen et al. | Short-term reservoir system operation for flood mitigation with 1D hydraulic model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |