Canal system gate scheduling optimization method based on distributed model predictive control and one-dimensional hydrodynamic model

Abstract

The invention discloses a canal system gate scheduling optimization method based on distributed model predictive control and a one-dimensional hydrodynamic model, and the method comprises the steps: collecting the hydrodynamic data, irrigation demands and canal system characteristics of a canal system, and taking the data as basic data; on the basis of the basic data, various disturbance conditions of the canal system are simulated, and corresponding hydrodynamic response data are obtained; constructing a one-dimensional hydrodynamic model based on the Saint-Venant equation; based on the one-dimensional hydrodynamic model and the hydrodynamic response data, constructing a distributed model prediction control architecture, and predicting the hydrodynamic data of the canal system; based on the prediction result, taking the minimum water level deviation and the minimum control cost as objective functions, and performing dynamic scheduling on the canal system; and performing feedback updating on a dynamic scheduling result of the channel system based on a dynamic rolling optimization mechanism. The method is suitable for multi-channel pool distributed water flow regulation and control, and particularly has important application value in efficient utilization of water resources in an irrigation area, flood scheduling and irrigation canal system optimization management.

Description

Technical Field

[0001] The present invention belongs to the technical field of water resources scheduling and management, and particularly relates to a method for optimizing the scheduling of canal gates based on distributed model predictive control and one-dimensional hydrodynamic model. Background Art

[0002] With the continuous development of water resources management technology and the upgrading of computer software and hardware infrastructure, modern water conservancy projects have put forward higher requirements for the efficient utilization of water resources. Channel flow regulation is an important link in the water resources management of irrigation areas, and its scientific scheduling is of great significance for improving irrigation efficiency, saving water resources, and ensuring agricultural production. Especially in multi-channel pond systems, the reasonable distribution and dynamic scheduling of water flow are the keys to maximizing the regional water use efficiency.

[0003] Traditional canal scheduling methods mainly rely on fixed water distribution schemes, usually based on surface water, ignoring the real-time changes of external environmental disturbances (such as rainfall, fluctuations in water demand, etc.), and unable to dynamically respond to the changes in actual situations. Although this method is relatively simple in engineering implementation, when encountering complex hydrological conditions (such as sudden rainfall or local water shortage), it is often difficult to adjust the water supply plan, which may lead to waste of water resources or inability to meet irrigation demands. In addition, in multi-channel pond systems, due to the lack of effective distributed control mechanisms and accurate dynamic prediction, the existing regulation methods are prone to unbalanced resource allocation between upstream and downstream channels, thus limiting the overall system operation efficiency.

[0004] In recent years, with the gradual maturity of hydrodynamic modeling technology, the one-dimensional hydrodynamic model based on the Saint-Venant equation has been widely used to simulate and analyze the dynamic changes of channel flow, providing a theoretical basis for canal scheduling. However, it is difficult to achieve real-time optimal control of complex systems only relying on the hydrodynamic model. Therefore, how to combine advanced control methods with accurate hydrodynamic models has become the key to solving the current channel regulation problem.

[0005] Distributed Model Predictive Control (DMPC), as an advanced optimal control technology, can divide complex systems into several subsystems, perform local optimization respectively, and achieve overall optimal control through global coordination. DMPC has strong distributed computing capabilities and real-time response capabilities, and can achieve efficient dynamic regulation of water flow in multi-channel pond systems. However, the current research on the combined application of DMPC and one-dimensional hydrodynamic models is still relatively scarce. Especially in multi-channel and branch canal systems, there is still a lack of mature methods on how to achieve coordinated control of multiple gates and coordinated optimization of upstream and downstream based on DMPC technology.

[0006] Therefore, there is an urgent need to propose a canal gate scheduling optimization method based on distributed model predictive control and one-dimensional hydrodynamic model to efficiently regulate complex canal systems through distributed model predictive control technology. Summary of the Invention

[0007] To solve the above technical problems, the present invention proposes a canal gate scheduling optimization method based on distributed model predictive control and one-dimensional hydrodynamic model. This method can respond in real time to changes in water level and flow rate caused by disturbances (such as gate opening adjustment, sudden rainfall, etc.), thereby improving water resource utilization efficiency and ensuring the accuracy and sustainability of canal system scheduling to solve the problems existing in the above prior art.

[0008] To achieve the above object, the present invention provides a canal gate scheduling optimization method based on distributed model predictive control and one-dimensional hydrodynamic model, including the following steps:

[0009] Collect hydrodynamic data, irrigation demand data, and canal system characteristic data of the canal system as basic data;

[0010] Based on the basic data, simulate various disturbance situations of the canal system to obtain corresponding hydrodynamic response data;

[0011] Based on the basic data, construct a one-dimensional hydrodynamic model based on the Saint-Venant equation;

[0012] Based on the one-dimensional hydrodynamic model and hydrodynamic response data, construct a distributed model predictive control framework to predict the hydrodynamic data of the canal system;

[0013] Based on the prediction results, take minimizing water level deviation and minimizing control cost as the optimization objective function to dynamically schedule the canal gates;

[0014] Based on the dynamic rolling optimization mechanism, feedback and update the canal gate scheduling results.

[0015] Optionally, the hydrodynamic data of the canal system includes: long-time series water level and flow rate data, historical change data of gate opening, and external disturbance data;

[0016] Irrigation demand data includes: water demand of each branch canal and controlled irrigation area;

[0017] Canal system characteristic data includes: channel length, canal top elevation, canal bottom elevation, slope coefficient, lining type, seepage coefficient, design flow rate, design water level, and canal system topology.

[0018] Optionally, the process of simulating various disturbance situations of the canal system based on the basic data to obtain corresponding hydrodynamic response data includes:

[0019] Build a channel model and set different scenarios of gate opening changes to simulate the dynamic impacts on the water levels and flows of each canal pond and branch canal respectively, and obtain the hydrodynamic response data of the canal system under various disturbance conditions.

[0020] Optionally, the Saint-Venant equations consist of a continuity equation and a momentum equation, and are used to describe the dynamic changes of water levels and flows in the canal system.

[0021] Optionally, the optimization objective function comprehensively combines several sub-objectives in the form of weighted summation, as shown in the following formula:

[0022]

[0023] where, J(u i ) represents the optimization objective of minimizing the system error and control energy consumption by optimizing the control variable ui. y i (k + i|k) represents the predicted state variable of subsystem i; Q i represents the weight matrix of the state variable; u i (k + j|k) represents the predicted control variable of subsystem i; R i represents the weight matrix of the control variable; p represents the prediction horizon; m represents the control horizon; k represents the current discrete time step; i represents the time index of state variable prediction, and its value range is 1 ≤ i ≤ p; j represents the time index of control variable optimization, and its value range is 1 ≤ j ≤ m.

[0024] Optionally, the constraint conditions of the optimization objective function include:

[0025] Hydraulic constraints:

[0026] Canal system design constraints:

[0027] Control input constraints:

[0028] where, E i is the coupling matrix, Z i is the coupling information from adjacent subsystems; n represents the index within the time step range, and its value range is 1 ≤ n ≤ p; Q i represents the actual flow of the channel; represents the designed flow of the channel; Z i represents the actual water level of the channel; represents the designed water level of the channel.

[0029] Optionally, based on the prediction results, the process of dynamically scheduling the canal gates with the optimization objective functions of minimizing water level deviation and minimizing control cost includes:

[0030] Based on the topological structure of the canal system, the canal system is divided into the main canal system and the branch canal system; based on the optimization objective function, the main canal system conducts overall flow scheduling and water level control for each canal pond in the main canal, and the branch canal system optimizes the gate opening degree and water distribution scheme inside each branch canal based on the scheduling result of the main canal system.

[0031] The present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory, and the processor executes the computer program to implement the steps of the method.

[0032] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method are implemented.

[0033] The present invention also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the method are implemented.

[0034] Compared with the prior art, the present invention has the following advantages and technical effects:

[0035] By combining the one-dimensional hydrodynamic model with the distributed model predictive control technology, the present invention effectively solves the problem of insufficient simulation of complex hydrodynamic processes in traditional canal system scheduling methods. The one-dimensional hydrodynamic model based on the Saint-Venant equation can accurately simulate the dynamic changes of water levels and flows in the canal system, capture the hydraulic coupling relationships between upstream and downstream canal ponds and between branch canals and the main canal, and provide accurate basic support for optimal regulation. At the same time, with its distributed optimization ability, the DMPC technology can achieve good coordination between global resource allocation and local fine control in the multi-canal pond and branch canal system, significantly improving the operating efficiency of the system.

[0036] By constructing an implicit multi-objective dynamic optimization model with the goals of minimizing water level deviation and minimizing control energy consumption, and adopting a "prediction-optimization-feedback" rolling control mechanism, the control strategy is continuously adjusted under the dynamic input of disturbance data, dynamically responding to the influence of gate opening degree changes or sudden disturbances, and ensuring the operating safety and water resource utilization efficiency of the channel system. During the optimization control process, by predicting future disturbance situations in advance and optimizing control decisions, the response ability of the system to emergencies (such as sudden rainfall or fluctuations in water use demand) can be significantly improved, reducing over-irrigation and water resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings constituting a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0038] Figure 1 It is a schematic flowchart of the method according to an embodiment of the present invention. Specific embodiments

[0039] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The following will describe this application in detail with reference to the drawings and in combination with the embodiments.

[0040] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0041] Embodiment 1

[0042] In this embodiment, a canal gate scheduling optimization method based on distributed model predictive control and one-dimensional hydrodynamic model is provided. This method realizes the efficient dynamic regulation of multiple canal pools and their branch canal systems by integrating the one-dimensional hydrodynamic model and the distributed model predictive control framework, and at the same time considers the real-time impact of gate operations on water levels and flows in complex canal systems.

[0043] In this embodiment, a one-dimensional hydrodynamic model based on the Saint-Venant equation is used to accurately simulate the dynamic changes of channel flows. Combining with distributed model predictive control (DMPC) technology, through optimizing the distribution of water levels and flows, the coordinated regulation of resources in the entire canal system is achieved. Each subsystem independently optimizes its local control objectives (such as gate openings and branch canal diversions), and the main system coordinates the resource allocation between subsystems through global optimization. Disturbance data (such as the impact of gate opening changes on water levels and flows) can be calculated offline through hydrodynamic simulation software (such as MIKE11) and used as important inputs for prediction and optimization in the DMPC controller.

[0044] In this embodiment, an implicit multi-objective optimization model that minimizes the water level deviation and minimizes the control energy consumption is constructed, and the optimal regulation strategy for the entire canal system is realized through the method of dynamic rolling prediction and control. On the premise of ensuring the safe operation of the canal and the agricultural water demand, it dynamically responds to disturbance changes, maximizes the utilization of water resources, reduces irrigation waste, and improves the agricultural economic benefits.

[0045] As Figure 1 shown, as a specific embodiment, the method includes the following steps:

[0046] Collect the hydrodynamic data, irrigation demand data, and canal system characteristic data of the canal system as basic data;

[0047] Based on the basic data, simulate various disturbance situations of the canal system to obtain the corresponding hydrodynamic response data;

[0048] Based on the above basic data, a one-dimensional hydrodynamic model based on the Saint-Venant equations is constructed;

[0049] Based on the one-dimensional hydrodynamic model and hydrodynamic response data, a distributed model predictive control architecture is constructed to predict the hydrodynamic data of the canal system;

[0050] Based on the prediction results, with the minimization of water level deviation and control cost as the optimization objective function, the canal system is dynamically scheduled;

[0051] Based on the dynamic rolling optimization mechanism, the dynamic scheduling results of the canal system are fed back and updated.

[0052] As an implementable approach, it includes the following steps:

[0053] S1: Basic data collection:

[0054] The construction of the model requires comprehensive collection of basic data such as hydrodynamics, irrigation demand, and canal system characteristics. These data will provide important support for the development of the one-dimensional hydrodynamic model and the distributed model predictive controller (DMPC). Hydrodynamic data includes long-term time series of water level and flow rate data, historical changes in gate openings, and external disturbances (such as short-term rainfall); in terms of irrigation demand, the water demand of each branch canal and the controlled irrigation area, etc. need to be collected; canal characteristic data includes canal length, canal top elevation, canal bottom elevation, slope coefficient, lining type, leakage coefficient, design flow rate, design water level, canal system topology, etc.

[0055] S2: Simulation and collection of disturbance data:

[0056] Based on the MIKE hydrodynamic simulation software, the simulation and collection of disturbance data for the canal system are carried out. By constructing a canal model and setting different gate opening change scenarios, the dynamic impacts on the water levels and flow rates of each canal pond and branch canal are simulated, and the hydrodynamic response data of the canal system under various disturbance conditions are obtained. These data reflect the actual impacts of gate operations on the hydraulic characteristics of the canal system and can truly reproduce the possible disturbance conditions during the operation of the canal.

[0057] After completing the disturbance simulation, the obtained water level and flow rate change data are used as disturbance inputs and transferred to the distributed model predictive control (DMPC) model to provide a basis for predicting future system states and optimizing scheduling decisions. In this way, accurate prediction and rolling optimization under disturbance conditions can be achieved, enabling the DMPC controller to dynamically adjust the canal system scheduling scheme and further enhancing the stability and response ability of the system. The accurate acquisition of the above disturbance data is a key link to ensure the operation effect of the model.

[0058] S3: Construction of the one-dimensional hydrodynamic model:

[0059] Based on the Saint-Venant equations, a one-dimensional hydrodynamic model is constructed to describe the dynamic changes of water levels and flows in a channel system. The Saint-Venant equations consist of a continuity equation and a momentum equation, and their expressions are as follows:

[0060]

[0061] This formula is the Continuity Equation in the Saint-Venant equations. In the formula, is the rate of change of the flow Q along the space S, which describes the change in the spatial distribution of the water flow. If the flow decreases within a certain section of space, this may mean that part of the water is stored within that section; is the rate of change of the cross-sectional area A with respect to time t, which describes the trend of the water flow accumulating or decreasing over time. The change in the cross-sectional area directly reflects the change in the water level. q represents the increase or decrease in the amount of water per unit length (source or sink term). If q > 0, it means that water is flowing in (such as rainfall or external water supply); if q < 0, it means that water is flowing out (such as evaporation or leakage).

[0062]

[0063] This formula is the Momentum Equation in the Saint-Venant equations, which describes the conservation of momentum of the water flow in a one-dimensional open channel flow and reflects the relationship between the momentum changes of the water flow in time and space. represents the rate of change of the flow Q with respect to time t, reflecting the momentum change of the flow at a certain cross-section due to the change over time. represents the rate of change of the momentum density of the flow Q in the space S, which reflects the momentum change of the flow due to the change in the flow velocity. The momentum correction coefficient α is used to correct the non-uniform distribution of the flow velocity across the cross-section, represents the influence of gravity g on the momentum of the water flow, which is specifically determined by the slope (the rate of change of the water surface or bottom elevation). If the slope is large, it will accelerate the water flow. represents the frictional loss between the water flow and the channel wall. The frictional force is usually proportional to the square of the flow velocity, denoted by |Q|. C is the Chezy coefficient, which can be replaced by the Manning coefficient to calculate the frictional force. AR combines the cross-sectional area and the hydraulic radius, reflecting the influence of the channel geometric characteristics on the frictional loss.

[0064] Meanwhile, according to the canal system topological structure and upstream and downstream boundary conditions, the flow balance and water level constraint relationships of each canal pond are defined. These relationships constitute the overall framework of the one-dimensional hydrodynamic model. The model can be further extended to a multi-canal pond system, and a dynamic hydrodynamic simulation model is formed by superimposing the state space equations of each canal pond.

[0065] This simulation model serves as the basis for the DMPC controller, providing the controller with real-time hydrodynamic prediction capabilities. By simulating the dynamic changes in water level and flow rate, the controller can accurately evaluate the impact of different gate operations on the system, thereby optimizing the scheduling decision. The establishment of the above model provides theoretical and technical support for the efficient regulation of the canal system.

[0066] S4: Design of the distributed simulation prediction controller:

[0067] In this embodiment, a distributed model predictive control (DMPC) architecture is adopted. By constructing a state space model and an optimization objective function, the dynamic regulation of the multi-canal pond system is realized. The DMPC controller dynamically adjusts the gate opening of each subsystem by predicting the future changes in water level and flow rate to optimize the operating performance of the entire canal system. The variables mainly include: state variables, including the water level and flow rate of the canal pond or branch canal; control variables, including the gate opening, which is used as a means to adjust the water resource allocation and is used to describe the dynamic state of the canal system.

[0068] Optimization objective: For each subsystem i, the optimization objective function combines multiple sub-objectives and is defined in the form of weighted summation as:

[0069]

[0070] In the formula, J(u i ) represents the optimization objective of optimizing the control variable ui to minimize the system error and control energy consumption. y i (k + i|k) represents the predicted state variable of subsystem i; Q i represents the weight matrix of the state variable; u i (k + j|k) represents the predicted control variable of subsystem i; R i represents the weight matrix of the control variable; p represents the prediction horizon, which means that the system predicts the state y i in the next p time steps in the future, usually set by the user to ensure the stability of the optimization calculation; m represents the control horizon, which means that the optimization variable u i is optimized in the next m time steps in the future. Usually, m ≤ p to balance the computational complexity and control performance; k represents the current discrete time step, that is, the current time index of the system; i represents the time index of the state variable prediction, and the value range is 1 ≤ i ≤ p; j represents the time index of the control variable optimization, and the value range is 1 ≤ j ≤ m.

[0071] The design of the objective function aims to: minimize the water level deviation so that the actual water level in the canal pool is close to the target water level; minimize the control cost and reduce the adjustment frequency or amplitude of the gate opening; and balance the overall system stability and water resource utilization efficiency through weight adjustment.

[0072] Constraints of the objective function:

[0073] Hydraulic constraint: This constraint describes the coupling relationship between subsystem i and adjacent subsystems. E i is the coupling matrix, and Z i is the coupling information from neighboring subsystems.

[0074] Canal system design constraint: Ensure that the water level and flow do not exceed the designed safety range.

[0075] Control input constraint: The upper and lower limits of the control variable (gate opening) can be added with the following constraints.

[0076] Among them, E i is the coupling matrix, and Z i is the coupling information from neighboring subsystems; n represents the index within the time step range, usually used for each time step within the prediction range of the control system, and the value range is 1 ≤ n ≤ p; Q i represents the actual flow rate of the channel; represents the designed flow rate of the channel; Z i represents the actual water level of the channel; represents the designed water level of the channel.

[0077] To achieve distributed optimization, the above problem is transformed into an augmented Lagrangian function:

[0078]

[0079] In the formula, λ is the Lagrange multiplier and ρ is the penalty parameter.

[0080] Distributed solution is achieved through the Alternating Direction Method of Multipliers (ADMM):

[0081] Update the control variable,

[0082] Update the coupling variable,

[0083] Update the dual variable,

[0084] According to the topological structure of the canal system, the system is divided into two levels: the main canal and the branch canals. Main canal system: Overall flow scheduling and water level control are carried out for each canal pond in the main canal. For the branch canal system, each branch canal is regarded as an independent subsystem, and the opening of its internal gates and the water distribution plan are optimized. Through the distributed optimization architecture, the main system and the subsystems work together to dynamically adjust the control strategy to achieve the global optimization of the operation performance of the canal system.

[0085] In this embodiment, efficient cooperation between the global and local levels is achieved through the master-slave control architecture. The main system is responsible for global optimization and resource allocation, while the subsystems independently optimize the local canal ponds. Through the coordinated operation of the master-slave architecture, not only the computational pressure on the main system is reduced, but also the real-time response ability of the subsystems to local disturbances is improved. This embodiment is applicable to complex canal systems in large and medium-sized irrigation areas, especially in scenarios of regulating multiple canal ponds and multiple branch canals, and has broad application value, providing important technical support for the improvement of future irrigation area water resource management and agricultural water use efficiency.

[0086] S5: Dynamic rolling optimization mechanism:

[0087] The dynamic rolling optimization mechanism is adopted. Through distributed model predictive control (DMPC), a closed-loop control of "prediction-optimization-feedback" is realized to ensure the high-efficiency adaptability of the canal scheduling system to the dynamic environment. The core of the rolling optimization mechanism lies in predicting and optimizing based on real-time disturbance data and feeding the optimization results back to the control strategy in real time, thereby dynamically adjusting the opening of the gates and optimizing the water resource allocation.

[0088] First, the system predicts the system state within the future prediction time domain according to the state space equation, the control variable sequence, and the known disturbance data. Then, with the control variables and disturbances as inputs and combined with the system state variables, a rolling optimization solution is carried out for the optimization objectives of minimizing the water level deviation and minimizing the control energy consumption. Finally, the optimization results are used to update the control variables (gate openings) at the current time step to achieve real-time response to disturbances. At the same time, the actually observed water level, flow rate and other data are used as feedback to update the model parameters and rollingly optimize the control strategy for the next time step.

[0089] Embodiment Two

[0090] This embodiment also provides a computer device, including a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the method.

[0091] Embodiment Three

[0092] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method are implemented.

[0093] Embodiment Four

[0094] This embodiment also provides a computer program product, including a computer program which, when executed by a processor, implements the steps of the method.

[0095] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the technical field of the present application within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An optimization method for canal gate scheduling based on distributed model predictive control and one-dimensional hydrodynamic model, characterized in that It includes the following steps: Collect the hydrodynamic data, irrigation demand data, and canal system characteristic data of the canal system as basic data; Based on the basic data, simulate various disturbance situations of the canal system to obtain corresponding hydrodynamic response data; Based on the basic data, construct a one-dimensional hydrodynamic model based on the Saint-Venant equations; Based on the one-dimensional hydrodynamic model and the hydrodynamic response data, construct a distributed model predictive control architecture to predict the hydrodynamic data of the canal system; Based on the prediction results, take minimizing the water level deviation and minimizing the control cost as the optimization objective function, and perform dynamic scheduling on the canal gates; Based on the dynamic rolling optimization mechanism, feedback and update the gate scheduling results of the canal system.

2. The method according to claim 1, wherein: The hydrodynamic data of the canal system includes: long-time series water level and flow data, historical change data of gate openings, and external disturbance data; The irrigation demand data includes: water demand of each branch canal and the controlled irrigation area; The canal system characteristic data includes: canal length, canal top elevation, canal bottom elevation, slope coefficient, lining type, seepage coefficient, design flow rate, design water level, and canal system topology.

3. The method according to claim 1, wherein: The process of simulating various disturbance situations of the canal system based on the basic data to obtain corresponding hydrodynamic response data includes: Construct a canal model, set different gate opening change scenarios, respectively simulate the dynamic impacts on the water levels and flows of each canal pool and branch canal, and obtain the hydrodynamic response data of the canal system under various disturbance situations.

4. The method according to claim 1, wherein: The Saint-Venant equations consist of a continuity equation and a momentum equation, and are used to describe the dynamic changes of water levels and flows in the canal system.

5. The method according to claim 1, wherein: The optimization objective function adopts a weighted summation form to synthesize several sub-objectives, as shown in the following formula: Among them, J(u i ) represents the optimization objective of minimizing the system error and control energy consumption by optimizing the control variable ui. y i (k + i|k) represents the predicted state variable of subsystem i; Q i represents the weight matrix of the state variable; u i (k + j|k) represents the predicted control variable of subsystem i; R i represents the weight matrix of the control variable; p represents the prediction horizon; m represents the control horizon; k represents the current discrete time step; i represents the time index of the state variable prediction, with a value range of 1 ≤ i ≤ p; j represents the time index of the control variable optimization, with a value range of 1 ≤ j ≤ m.

6. The method according to claim 5, wherein: The constraint conditions of the optimization objective function include: Hydraulic constraint: Canal system design constraints: Control input constraint: Among them, E i is the coupling matrix, Z i is the coupling information from adjacent subsystems, n represents the index within the time step range, and the value range is 1 ≤ n ≤ p; Q i represents the actual flow rate of the channel; represents the designed flow rate of the channel; Z i represents the actual water level of the channel; represents the designed water level of the channel.

7. The method according to claim 1, wherein: The process of performing dynamic scheduling on the canal gates based on the prediction results with minimizing the water level deviation and minimizing the control cost as the optimization objective function includes: Based on the topology of the canal system, divide the canal system into a main canal system and a branch canal system; based on the optimization objective function, the main canal system performs overall flow scheduling and water level control on each canal pool of the main canal, and the branch canal system optimizes the gate openings and water distribution schemes inside each branch canal based on the scheduling results of the main canal system.

8. A computer device, comprising a memory, a processor, and a computer program stored on the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1-7.

Images