Water conservancy facility intelligent scheduling method based on Levy flight optimization and BIM

Abstract

The invention discloses a water conservancy facility intelligent scheduling method based on Levy flight optimization and BIM. The method comprises the following steps: S1, constructing a full-life-cycle digital three-dimensional model of a target water conservancy facility on a BIM platform; s2, collecting a real-time monitoring data set of the target water conservancy facility, and performing data mapping on the collected real-time monitoring data set and the full-life-cycle digital three-dimensional model; s3, forming an optimization framework of the scheduling problem; s4, initializing control parameters of the Levy flight optimization algorithm; s5, performing global search on the scheduling problem by utilizing a Levy flight optimization algorithm, and generating a plurality of candidate scheduling schemes in the optimization model by adopting a random jump and adaptive adjustment mechanism; and S6, selecting an optimal scheduling scheme, mapping the selected optimal scheduling scheme to a control system of the target water conservancy facility, and generating a corresponding scheduling control instruction. According to the invention, the intelligence, adaptability and operability of the water conservancy facility scheduling system are greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of water conservancy technology, and in particular to an intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM. Background Art

[0002] With the continuous development of digital twins and information technology in the infrastructure field, water conservancy facility management is gradually moving towards intelligent and digital transformation. As a core component of the national water resources allocation, flood control and disaster reduction, and irrigation and water supply systems, the scientific and real-time scheduling and management of water conservancy facilities are directly related to regional water security and operational efficiency.

[0003] In existing technologies, water conservancy facility scheduling generally relies on rule bases, expert experience, and static mathematical models for decision-making. Although such methods have certain adaptability under specific working conditions, they are difficult to cope with the highly dynamic and uncertain nature of current environmental conditions. Specifically: First, traditional scheduling models are usually established under static assumptions and lack a real-time status update mechanism during facility operation, making it difficult to respond quickly to sudden floods, droughts, and equipment failures; second, the structural constraints, operating logic, and equipment semantic information of water conservancy facilities are often not fully integrated in scheduling decisions, making it difficult to implement scheduling plans; third, scheduling optimization algorithms are mostly based on local search strategies or linear programming methods. When dealing with high-dimensional, nonlinear, and multi-constrained problems, they are prone to falling into local optimality and it is difficult to obtain a global solution to the scheduling problem.

[0004] On the other hand, although Building Information Modeling (BIM) has been promoted in the design, construction, and management of water conservancy projects in recent years, its deep integration into the scheduling decision-making level remains relatively weak. Currently, BIM is mainly used for visualization and structural collaborative management, and its massive amount of operational data and component semantic information has not been effectively utilized for the intelligent generation and optimization of scheduling strategies. In addition, existing scheduling systems often lack a highly coupled data channel with BIM models, resulting in data update delays and information fragmentation, further restricting the real-time and scientific nature of scheduling solutions.

[0005] Therefore, it is urgent to propose a new method for intelligent scheduling of water conservancy facilities that integrates BIM technology and intelligent optimization algorithms, and has structural constraint adaptability, real-time response capability and global search capability, so as to effectively solve the above technical bottlenecks. Summary of the Invention

[0006] One purpose of the present invention is to propose an intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM. The present invention greatly improves the intelligence, adaptability and operability of the water conservancy facility scheduling system.

[0007] According to an embodiment of the present invention, a method for intelligent scheduling of water conservancy facilities based on Lévy flight optimization and BIM includes the following steps:

[0008] S1. Build a digital 3D model of the entire life cycle of the target water conservancy facility on the BIM platform;

[0009] S2. Collect real-time monitoring data sets of target water conservancy facilities and map them to the digital 3D model of the entire life cycle to achieve dynamic updates of the digital 3D model data of the entire life cycle;

[0010] S3. Based on the real-time, dynamically updated data on the BIM platform, extract the current operating status information and scheduling boundary conditions of the target water conservancy facilities, and establish an optimization model for the water conservancy facility scheduling problem, forming an optimization framework for the scheduling problem;

[0011] S4. Initialize the control parameters of the Lévy flight optimization algorithm and generate an initial solution group based on the scheduling problem optimization model extracted from the BIM platform. At the same time, introduce an adaptive step size adjustment mechanism based on BIM semantic constraints.

[0012] S5. Use the Lévy flight optimization algorithm to perform a global search for the scheduling problem, employ random jumps and adaptive adjustment mechanisms to generate multiple candidate scheduling solutions within the optimization model, and map and verify each candidate scheduling solution through the BIM platform.

[0013] S6. Comprehensively evaluate the candidate scheduling plans based on the preset scheduling objectives, select the optimal scheduling plan, and map the selected optimal scheduling plan to the control system of the target water conservancy facility to generate corresponding scheduling control instructions.

[0014] Optionally, the S1 includes the following steps:

[0015] S11. Obtain an engineering design dataset of the target water conservancy facility, including structural component numbers, geometric dimensions, material types, and spatial position parameters, and use the engineering design dataset as the structural basis for constructing a full life cycle digital 3D model to construct a structural information matrix;

[0016] S12. Connect to the real-time operation monitoring system of the target water conservancy facility and collect historical and current operation status data sets, including water level monitoring values, flow records, pump station status, gate opening and closing status, and equipment start and stop time operation parameters, to construct an operation status information matrix;

[0017] S13. Based on the real-time environmental perception system in the water conservancy facility operation area, collect environmental parameter data sets, including rainfall intensity, evaporation, temperature, wind speed, and soil moisture, which are external environmental information affecting the operation of water conservancy facilities, and construct an environmental parameter matrix;

[0018] S14. Extract a set of key scheduling parameters, including the facility operating capacity upper limit, scheduling time window, response delay limit, and operation frequency constraint, and convert the key scheduling parameter constraint vector;

[0019] S15. The structural information matrix M structure , operating status information matrix M status , environmental parameter matrix M env And the key scheduling parameter constraint vector P key Integrate into the BIM platform to build a digital 3D model of the entire life cycle of the target water conservancy facilities BIM

[0020] M BIM = <M structure ,M status ,M env ,P key >.

[0021] Optionally, the S2 includes the following steps:

[0022] S21. Real-time monitoring system deployed based on water conservancy facilities collects real-time monitoring data streams Each piece of data d t Including the current water level h t , flow rate q t , device status t , valve status v t And the environmental parameter set E t , construct real-time monitoring data vector:

[0023] d t =(h t ,q t ,s t ,v t ,E t );

[0024] S22. The real-time monitoring data stream D real and the structural information matrix M structure and the operating status information matrix M status Perform time tag alignment and map the acquisition time t to the full life cycle digital 3D model M in the BIM platform. BIM The corresponding running node builds the time-state mapping function:

[0025] f map :d t →M BIM (t);

[0026] Among them, f mapRepresents the real-time data mapping function, which enables each set of real-time monitoring data to be injected into the corresponding scheduling unit node;

[0027] S23. Operation status information matrix M in the BIM platform status Perform dynamic updates and update the new monitoring data vector d according to the set time update cycle t Added updated operational status information matrix:

[0028]

[0029] in, represents the operating status information matrix before time t, Indicates updated status information;

[0030] S24. Update the running status information matrix and the environmental parameter matrix M env The current window data E t Jointly inject into the BIM platform to dynamically update the digital 3D model of the entire life cycle

[0031]

[0032] Optionally, S3 includes the following steps:

[0033] S31. Dynamically updated full lifecycle digital 3D model based on the BIM platform Extract the current operating status information of the target water conservancy facility and define the current operating status information vector. The current operating status information vector includes the current water level status, the current flow status, the current equipment operating status, the current valve opening and closing status, and the current environmental parameter set.

[0034] S32. Based on the key scheduling parameter constraint vector P key , extract the current scheduling boundary conditions, including the scheduling capacity upper limit of the target water conservancy facilities, the allowed scheduling time window, the response delay limit and the operation frequency constraint, and construct the scheduling constraint condition set C boundary ;

[0035] S33. According to the running status information vector X t+Δt and the scheduling constraint set C boundary , an optimization model for water conservancy facility scheduling is established. The optimization model defines the state space of the scheduling scheme as:

[0036]

[0037] Among them, S stateRepresents all schedulable states of the target water conservancy facilities under current constraints;

[0038] S34.Based on the full life cycle digital 3D model in the BIM platform The structure and operation constraints of define the feasible solution space:

[0039] Satisfy C boundary And in Valid within};

[0040] in, is the i-th feasible scheduling solution, and U is the set of all possible scheduling actions;

[0041] S35. Based on the actual requirements of intelligent scheduling of target water conservancy facilities, define the scheduling performance objective function of the optimization model:

[0042]

[0043] Among them, f level is the water level control error function, which reflects the degree to which the scheduling scheme satisfies the water level control target, f flow is the flow optimization control function, which reflects the adaptability of the scheduling scheme to flow demand, f delay is the response delay penalty function, which reflects the real-time response capability of the scheduling scheme, and w1, w2, and w3 are the weight coefficients of the scheduling target;

[0044] S36. Comprehensive dispatchable state S state , feasible solution space S solution and the scheduling performance objective function f opt , and build a complete optimization framework for the intelligent scheduling of water conservancy facilities.

[0045] Optionally, the S4 includes the following steps:

[0046] S41. Full life cycle digital 3D model based on BIM platform Analyze the criticality and operational sensitivity of each dispatching node of the target water conservancy facility, and dynamically calculate the sensitivity weight coefficient η of each node based on the node's physical location, importance level and environmental impact. i ,Define the node sensitivity weight vector, and a high value of the sensitive weight coefficient indicates that the node has a greater impact on the overall safe operation and scheduling efficiency of the facility;

[0047] S42. Initialize the control parameter set of the Lévy flight optimization algorithm, which includes the basic jump step factor α0, the Lévy distribution stability index β, the maximum number of iterations T max and population size N p , construct the initial control parameter set Θ LFO;

[0048] S43. Feasible solution space S according to the optimization model solution With the state space S state , generate the initial candidate solution group:

[0049]

[0050] Among them, each initial candidate solution It is a set of feasible scheduling strategies that meet the physical structure and operation boundary conditions of water conservancy facilities;

[0051] S44. Introducing the sensitive weight coefficient η i and environmental change rateζ t Adaptive step adjustment mechanism, environmental change rate ζ t The algorithm search strategy is adjusted in real time through weighted calculation of the water level fluctuation rate, flow change trend and the impact of meteorological emergencies in real-time monitoring data. The adaptive step size is defined as:

[0052] α i (t) = α0·(1+η i ·ζ t ).

[0053] Optionally, the S5 includes the following steps:

[0054] S51. In each round of iteration, based on the current candidate scheduling solution group X k And the corresponding adaptive step size α i (t), perform Lévy flying jump operation to generate the next generation candidate scheduling solution group X k+1 :

[0055]

[0056] in, It represents the i-th scheduling plan in the k-th iteration. The scheduling plan consists of a set of adjustable points and their operating states, representing a feasible scheduling path for the target water conservancy facility. represents the i-th candidate scheduling solution in the k+1-th iteration, x best represents the current global optimal scheduling solution, α i (t) is the adaptive jump step of the i-th node at time t, L β is a random variable that follows the Lévy distribution;

[0057] S52. For each candidate scheduling solution generated Calculate its scheduling performance objective function

[0058] S53. Calling the full life cycle digital 3D model in the BIM platform For each candidate scheduling solution Perform physical structure constraint mapping verification:

[0059]

[0060] Among them, the function f map (·) Determine whether the candidate solution is physically executable, including whether there are structural conflicts, path redundancies, or scheduling logic violations. A return value of True indicates that the solution structure is legal.

[0061] S54. For candidate scheduling solutions whose mapping verification is False, call the structural constraint correction mechanism to perform local optimization:

[0062]

[0063] Among them, ∈ i is the perturbation coefficient, γ i is the node semantic weight factor, combined with the scheduling sensitivity setting of the node in the BIM model, and the key nodes are adjusted first. is the gradient direction of the objective function of the candidate scheduling solution in the local space;

[0064] S55. All candidate scheduling solutions after mapping verification and local correction Recalculate its scheduling performance objective function value and compare it with the current global optimal scheduling solution x best Compare, if satisfied Then the updated global optimal scheduling solution is

[0065] S56. Repeat S51 to S55 until the maximum number of iterations T is reached max , and finally obtain a set of scheduling solutions that meet the physical structure, real-time operating status and scheduling objectives of water conservancy facilities.

[0066] Optionally, the S6 includes the following steps:

[0067] S61. Based on the scheduling performance objective function For the candidate scheduling solution group X k+1 Comprehensively evaluate all legal scheduling schemes in the structure, and select the global optimal scheduling scheme based on the following strategy If all candidate solutions meet the structural constraints and scheduling boundary conditions of the water conservancy facilities, the solution with the smallest objective function value is selected:

[0068]

[0069] If there are multiple solutions with similar objective function values, that is, the difference is less than the set threshold, they will be further sorted according to the equipment scheduling frequency, energy consumption cost and operation complexity, and the solution with low scheduling operation frequency and simple control action set will be given priority;

[0070] If operating in a high-frequency emergency environment, a real-time emergency factor θ is introduced t Dynamically modify the weight of the objective function, recalculate, and select a more responsive scheduling solution;

[0071] S62. Select the optimal scheduling solution Mapped to the digital 3D model of the entire life cycle of the target water conservancy facility Each scheduling control node in the BIM model generates a control instruction set U based on the semantic label and physical connection relationship of each control node in the BIM model. ctrl ;

[0072] S63. Set the control instruction set U ctrl It is sent to the control system of the target water conservancy facility to implement the selected optimal scheduling plan, start the operation status monitoring mechanism, and record the scheduling result data for subsequent model optimization and scheduling feedback adjustment.

[0073] Optional, structural expression of the optimal scheduling solution:

[0074] If the water level warning value h t+Δt >h safe , and the traffic prediction value q t+Δt >q thresh , then the emergency control plan is triggered and a dispatching path with rapid flood discharge capability is selected;

[0075] If the water level fluctuation rate If the energy consumption is low, the environmental parameters are stable, and the rainfall in the next 2 hours is predicted to be less than the set threshold, the scheduling plan with the lowest energy consumption is selected;

[0076] If there are multiple concurrent tasks in the target area, the flow allocation priority rule is adopted to meet the downstream irrigation priority scheduling conditions;

[0077] If the current device operating frequency exceeds the set maximum operating frequency F op , then the candidate solutions containing excessive operations are eliminated, and the path solutions with high structural stability are prioritized.

[0078] The beneficial effects of the present invention are:

[0079] (1) The present invention constructs a node sensitivity weight vector and combines it with the real-time environmental change rate to form an adaptive step size, thereby realizing a dynamic amplified jump search for highly sensitive areas of hydrological emergencies and making fine-tuning corrections to the structural stable areas, thereby improving the robustness and response speed of the optimization process. In flood control scheduling simulations, this mechanism can improve the efficiency of obtaining the global optimal solution and shorten the average response time.

[0080] (2) The present invention utilizes the type labels, operating status and semantic logic of components in the BIM model to construct a scheduling semantic verification function, performs real-time structural mapping verification on each generation of candidate scheduling schemes, and fine-tunes and corrects illegal scheduling schemes through a semantically guided local gradient perturbation mechanism to ensure that all final output scheduling solutions have structural executableness.

[0081] (3) The present invention constructs state space and feasible solution space based on the structural information, operating status and scheduling boundaries dynamically updated by the BIM platform, and designs a multi-objective function that integrates water level control accuracy, flow optimization objectives and response timeliness. During the optimization process, the flexible switching and coordinated balance between scheduling tasks are achieved through dynamic adjustment of weights. The multi-layer collaborative optimization mechanism is particularly suitable for intelligent scheduling applications of water conservancy facilities in different operation and maintenance stages and functional modes, and effectively supports comprehensive scheduling decisions under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0083] Figure 1 This is a flow chart of the intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM proposed in the present invention. DETAILED DESCRIPTION

[0084] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0085] refer to Figure 1 , an intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM, including the following steps:

[0086] S1. Build a digital 3D model of the entire life cycle of the target water conservancy facility on the BIM platform;

[0087] S2. Collect real-time monitoring data sets of target water conservancy facilities and map them to the digital 3D model of the entire life cycle to achieve dynamic updates of the digital 3D model data of the entire life cycle;

[0088] S3. Based on the real-time, dynamically updated data on the BIM platform, extract the current operating status information and scheduling boundary conditions of the target water conservancy facilities, and establish an optimization model for the water conservancy facility scheduling problem, forming an optimization framework for the scheduling problem;

[0089] S4. Initialize the control parameters of the Lévy flight optimization algorithm and generate an initial solution group based on the scheduling problem optimization model extracted from the BIM platform. At the same time, introduce an adaptive step size adjustment mechanism based on BIM semantic constraints.

[0090] S5. Use the Lévy flight optimization algorithm to perform a global search for the scheduling problem, employ random jumps and adaptive adjustment mechanisms to generate multiple candidate scheduling solutions within the optimization model, and map and verify each candidate scheduling solution through the BIM platform.

[0091] S6. Comprehensively evaluate the candidate scheduling plans based on the preset scheduling objectives, select the optimal scheduling plan, and map the selected optimal scheduling plan to the control system of the target water conservancy facility to generate corresponding scheduling control instructions.

[0092] In this embodiment, S1 includes the following steps:

[0093] S11. Obtain an engineering design dataset for the target water conservancy facility. The engineering design dataset includes structural component numbers, geometric dimensions, material types, and spatial position parameters. Use the engineering design dataset as the structural foundation for constructing a full-lifecycle digital 3D model and building a structural information matrix.

[0094] S12. Connect to the real-time operation monitoring system of the target water conservancy facility and collect historical and current operation status data sets, including water level monitoring values, flow records, pump station status, gate opening and closing status, and equipment start and stop time operation parameters, to construct an operation status information matrix;

[0095] S13. Based on the real-time environmental perception system in the water conservancy facility operation area, collect environmental parameter data sets, including rainfall intensity, evaporation, temperature, wind speed, and soil moisture, which are external environmental information affecting the operation of water conservancy facilities, and construct an environmental parameter matrix;

[0096] S14. Extract a set of key scheduling parameters, including facility operating capacity upper limit, scheduling time window, response delay limit, and operation frequency constraint, and convert the key scheduling parameter constraint vector;

[0097] S15. The structural information matrix M structure , operating status information matrix M status , environmental parameter matrix M env And the key scheduling parameter constraint vector P keyIntegrate into the BIM platform to build a digital 3D model of the entire life cycle of the target water conservancy facilities BIM

[0098] M BIM = <M structure ,M status ,M env ,P key >.

[0099] In this embodiment, S2 includes the following steps:

[0100] S21. Real-time monitoring system deployed based on water conservancy facilities collects real-time monitoring data streams Each piece of data d t Including the current water level h t , flow rate q t , device status t , valve status v t And the environmental parameter set E t , construct real-time monitoring data vector:

[0101] d t =(h t ,q t ,s t ,v t ,E t );

[0102] S22. Real-time monitoring of data flow D real and the structural information matrix M structure and the operating status information matrix M status Perform time tag alignment and map the acquisition time t to the full life cycle digital 3D model M in the BIM platform. BIM The corresponding running node builds the time-state mapping function:

[0103] f map :d t →M BIM (t);

[0104] Among them, f map Represents the real-time data mapping function, which enables each set of real-time monitoring data to be injected into the corresponding scheduling unit node;

[0105] S23. Operation status information matrix M in the BIM platform status Perform dynamic updates and update the new monitoring data vector d according to the set time update cycle t Added updated operational status information matrix:

[0106]

[0107] in, represents the operating status information matrix before time t, Indicates updated status information;

[0108] S24. Update the running status information matrix and the environmental parameter matrix M env The current window data E t Jointly inject into the BIM platform to dynamically update the digital 3D model of the entire life cycle

[0109]

[0110] In this embodiment, S3 includes the following steps:

[0111] S31. Dynamically updated full lifecycle digital 3D model based on the BIM platform Extract the current operating status information of the target water conservancy facility and define the current operating status information vector. The current operating status information vector includes the current water level status, the current flow status, the current equipment operating status, the current valve opening and closing status, and the current environmental parameter set.

[0112] S32. Based on the key scheduling parameter constraint vector P key , extract the current scheduling boundary conditions, including the scheduling capacity upper limit of the target water conservancy facilities, the allowed scheduling time window, the response delay limit and the operation frequency constraint, and construct the scheduling constraint condition set C boundary ;

[0113] S33. According to the running status information vector X t+Δt and the scheduling constraint set C boundary , an optimization model for water conservancy facility scheduling is established. The optimization model defines the state space of the scheduling scheme as:

[0114]

[0115] Among them, S state Represents all schedulable states of the target water conservancy facilities under current constraints;

[0116] S34.Based on the full life cycle digital 3D model in the BIM platform The structure and operation constraints of define the feasible solution space:

[0117]

[0118] in, is the i-th feasible scheduling solution, and U is the set of all possible scheduling actions;

[0119] S35. Based on the actual requirements of intelligent scheduling of target water conservancy facilities, define the scheduling performance objective function of the optimization model:

[0120]

[0121] Among them, f level is the water level control error function, which reflects the degree to which the scheduling scheme satisfies the water level control target, f flow is the flow optimization control function, which reflects the adaptability of the scheduling scheme to flow demand, f delay is the response delay penalty function, which reflects the real-time response capability of the scheduling scheme, and w1, w2, and w3 are the weight coefficients of the scheduling target;

[0122] S36. Comprehensive dispatchable state S state , feasible solution space S solution and the scheduling performance objective function f opt , and build a complete optimization framework for the intelligent scheduling of water conservancy facilities.

[0123] In this embodiment, S4 includes the following steps:

[0124] S41. Full life cycle digital 3D model based on BIM platform Analyze the criticality and operational sensitivity of each dispatching node of the target water conservancy facility, and dynamically calculate the sensitivity weight coefficient η of each node based on the node's physical location, importance level and environmental impact. i ,Define the node sensitivity weight vector, and a high value of the sensitive weight coefficient indicates that the node has a greater impact on the overall safe operation and scheduling efficiency of the facility;

[0125] S42. Initialize the control parameter set of the Lévy flight optimization algorithm, which includes the basic jump step factor α0, the Lévy distribution stability index β, the maximum number of iterations T max and population size N p , construct the initial control parameter set Θ LFO ;

[0126] S43. Feasible solution space S according to the optimization model solution With the state space S state , generate the initial candidate solution group:

[0127]

[0128] Among them, each initial candidate solution It is a set of feasible scheduling strategies that meet the physical structure and operation boundary conditions of water conservancy facilities;

[0129] S44. Introducing the sensitive weight coefficient η iand environmental change rateζ t Adaptive step adjustment mechanism, environmental change rate ζ t The algorithm search strategy is adjusted in real time through weighted calculation of the water level fluctuation rate, flow change trend and the impact of meteorological emergencies in real-time monitoring data. The adaptive step size is defined as:

[0130] α i (t) = α0·(1+η i ·ζ t ).

[0131] In this embodiment, S5 includes the following steps:

[0132] S51. In each round of iteration, based on the current candidate scheduling solution group X k And the corresponding adaptive step size α i (t), perform Lévy flying jump operation to generate the next generation candidate scheduling solution group X k+1 :

[0133]

[0134] in, It represents the i-th scheduling plan in the k-th iteration. The scheduling plan consists of a set of adjustable points and their operating states, representing a feasible scheduling path for the target water conservancy facility. represents the i-th candidate scheduling solution in the k+1-th iteration, x best represents the current global optimal scheduling solution, α i (t) is the adaptive jump step of the i-th node at time t, L β is a random variable that follows the Lévy distribution;

[0135] S52. For each candidate scheduling solution generated Calculate its scheduling performance objective function

[0136] S53. Calling the full life cycle digital 3D model in the BIM platform For each candidate scheduling solution Perform physical structure constraint mapping verification:

[0137]

[0138] Among them, the function f map (·) Determine whether the candidate solution is physically executable, including whether there are structural conflicts, path redundancies, or scheduling logic violations. A return value of True indicates that the solution structure is legal.

[0139] S54. For candidate scheduling solutions whose mapping verification is False, call the structural constraint correction mechanism to perform local optimization:

[0140]

[0141] Among them, ∈ i is the perturbation coefficient, γ i is the node semantic weight factor, combined with the scheduling sensitivity setting of the node in the BIM model, and the key nodes are adjusted first. is the gradient direction of the objective function of the candidate scheduling solution in the local space;

[0142] S55. All candidate scheduling solutions after mapping verification and local correction Recalculate its scheduling performance objective function value and compare it with the current global optimal scheduling solution x best Compare, if satisfied Then the updated global optimal scheduling solution is

[0143] S56. Repeat S51 to S55 until the maximum number of iterations T is reached max , and finally obtain a set of scheduling solutions that meet the physical structure, real-time operating status and scheduling objectives of water conservancy facilities.

[0144] In this embodiment, S6 includes the following steps:

[0145] S61. Based on the scheduling performance objective function For the candidate scheduling solution group X k+1 Comprehensively evaluate all legal scheduling schemes in the structure, and select the global optimal scheduling scheme based on the following strategy If all candidate solutions meet the structural constraints and scheduling boundary conditions of the water conservancy facilities, the solution with the smallest objective function value is selected:

[0146]

[0147] If there are multiple solutions with similar objective function values, that is, the difference is less than the set threshold, they will be further sorted according to the equipment scheduling frequency, energy consumption cost and operation complexity, and the solution with low scheduling operation frequency and simple control action set will be given priority;

[0148] If operating in a high-frequency emergency environment, a real-time emergency factor θ is introduced t Dynamically modify the weight of the objective function, recalculate, and select a more responsive scheduling solution;

[0149] S62. Select the optimal scheduling solution Mapped to the digital 3D model of the entire life cycle of the target water conservancy facility Each scheduling control node in the BIM model generates a control instruction set U based on the semantic label and physical connection relationship of each control node in the BIM model. ctrl ;

[0150] S63. Set the control instruction set U ctrl It is sent to the control system of the target water conservancy facility to implement the selected optimal scheduling plan, start the operation status monitoring mechanism, and record the scheduling result data for subsequent model optimization and scheduling feedback adjustment.

[0151] In this implementation, the optimal scheduling solution is expressed as follows:

[0152] If the water level warning value h t+Δt >h safe , and the traffic prediction value q t+Δt >q thresh , then the emergency control plan is triggered and a dispatching path with rapid flood discharge capability is selected;

[0153] If the water level fluctuation rate If the energy consumption is low, the environmental parameters are stable, and the rainfall in the next 2 hours is predicted to be less than the set threshold, the scheduling plan with the lowest energy consumption is selected;

[0154] If there are multiple concurrent tasks in the target area, the flow allocation priority rule is adopted to meet the downstream irrigation priority scheduling conditions;

[0155] If the current device operating frequency exceeds the set maximum operating frequency F op , then the candidate solutions containing excessive operations are eliminated, and the path solutions with high structural stability are prioritized.

[0156] Example 1:

[0157] This embodiment uses the flood control scenario of the X River Basin in a certain city in XX Province as the background to illustrate the feasibility and effectiveness of the present invention under actual complex hydrological conditions.

[0158] The river basin covers 2 medium-sized reservoirs, 4 gate control points and 1 urban drainage station, and is mainly responsible for regional flood scheduling, urban waterlogging prevention and control, and agricultural irrigation water replenishment. In recent years, due to the frequent occurrence of extreme weather, the river basin has frequently experienced heavy rainfall, sudden increase in upstream water, and urban waterlogging. The combined effects have caused a series of operational difficulties such as delayed scheduling response, high intensity of manual operation, and scheduling schemes that cannot adapt to real-time changes. In order to cope with a typical heavy rainfall warning scenario in June 2024, the regional water conservancy bureau decided to adopt the scheduling system proposed in this invention for pilot application to replace the traditional scheduling mechanism based on static rule base and manual experience, and explore the deep integration scheduling strategy of intelligent optimization and digital model.

[0159] First, a complete 3D digital model of the entire life cycle of the region's water conservancy facilities was established using the BIM platform:

[0160] Structural data of the two reservoirs (dam height 35 meters, storage capacity 54 million cubic meters, two flood discharge gates);

[0161] Operating parameters of the four gates (gate width 6 meters, height 4 meters, electric opening and closing control);

[0162] The equipment status of a drainage station (three pumps with a total flow capacity of 9,000 cubic meters per hour);

[0163] Information on surrounding terrain, hydrological monitoring points, meteorological stations, and underground drainage networks;

[0164] Real-time data interface access: water level, pump operation, rainfall, wind speed, flow rate;

[0165] Historical operating data covers the past three years (January 2021 to April 2024);

[0166] At 12:00 noon on June 7, 2024, the meteorological department predicted that the basin would experience regional heavy rainfall in the next 24 hours, with an estimated rainfall of 210 mm, which falls under the Level III rainstorm warning standard. The model automatically triggered the intelligent scheduling module. At this time, the system accessed real-time monitoring data and detected the following parameters:

[0167] The upstream water level rose to 28.3 meters (the warning water level is 29 meters);

[0168] One water pump has been turned on at the downstream urban drainage station, with a flow rate of 2.1m 3 / s;

[0169] The current weather forecast shows that the rain will peak within 3 hours and last for 8 hours;

[0170] During the data preparation stage, the platform updates the BIM model based on the monitoring data, completes the joint modeling of the current structural status, operating status and environmental status of the water conservancy facilities, and automatically extracts scheduling constraints. The permitted flood discharge period is 14:00-02:00, the frequency of reservoir water release is limited to no more than once every 4 hours, and the continuous working time of a single pump at the drainage station shall not exceed 3 hours.

[0171] Subsequently, the system starts the Lévy flight optimization module to perform a global search of the scheduling state space and candidate solutions. The node sensitivity weights introduced in this process are as follows:

[0172] The sensitivity weight of the upstream reservoir node is 0.85 (because it directly affects the downstream water level changes);

[0173] The sensitivity weight of the urban drainage pump station node is 0.78 (due to the high impact of urban waterlogging);

[0174] The gate node sensitivity weights vary between 0.6–0.72;

[0175] Combined with the real-time water level change rate (it rose by 0.5 meters in the past hour), the system dynamically adjusted and optimized the jump step size, increased the search intensity of high-sensitivity nodes, generated 500 sets of feasible scheduling strategies in the initial population, and converged to the optimal scheduling solution after 400 iterations.

[0176] The optimal scheduling strategy results are as follows:

[0177] At 14:30, the first flood discharge gate of the upstream reservoir was opened, discharging 800m 3 / s;

[0178] 15:00-15:30 Start the second water pump of the urban drainage station (total flow 6000m 3 / h);

[0179] 16:30–17:30: Downstream gate #3 is opened to 60% to ensure safe drainage of agricultural irrigation areas;

[0180] At 21:00, the reservoir gates were closed and the water storage mode was switched to cope with subsequent continuous rainfall;

[0181] The dispatching instruction set has been verified through BIM model mapping and meets all equipment operating conditions and control logic. The instructions are directly issued to the reservoir dispatching control system and drainage station PLC equipment.

[0182] To verify the performance difference between the proposed method and the traditional method, a three-day simulation comparison test was conducted. The same initial boundary conditions were set, and the scheduling goal was to minimize the peak water level and shorten the duration of water accumulation. The results are as follows:

[0183]

[0184] In addition, sample analysis shows that the method of the present invention has a more obvious optimization effect on scenarios with complex structures and high scheduling logic coupling. In the past 10 hydrological event samples (collected from rainstorm and flood cases in 2022-2024), the scheduling system of the present invention can reduce the scheduling failure rate by an average of 62%, and shorten the operation and debugging time by about 41%, significantly improving the system's intelligent response capability and scheduling effectiveness.

[0185] In summary, this embodiment comprehensively demonstrates the operational flow, algorithm response process, and execution effect of the present invention in actual intelligent scheduling scenarios of water conservancy facilities, and fully verifies the superiority of this method over existing technologies in terms of structural constraint adaptability, scheduling real-time performance, and execution feasibility.

[0186] The present invention constructs a node sensitivity weight vector and combines it with the real-time environmental change rate to form an adaptive step size, thereby realizing a dynamic amplified jump search for highly sensitive areas of hydrological emergencies (heavy rainfall, water level surge), and making fine-tuning corrections to the structural stable areas, thereby improving the robustness and response speed of the optimization process. In flood control scheduling simulations, this mechanism can improve the efficiency of obtaining the global optimal solution and shorten the average response time.

[0187] The present invention utilizes the type labels, operating status and semantic logic of components in the BIM model to construct a scheduling semantic verification function, performs real-time structural mapping verification on each generation of candidate scheduling schemes, and fine-tunes and corrects illegal scheduling schemes through a semantically guided local gradient perturbation mechanism to ensure that all final output scheduling solutions have structural executable.

[0188] The present invention constructs state space and feasible solution space based on the structural information, operating status and scheduling boundaries dynamically updated by the BIM platform, and designs a multi-objective function that integrates water level control accuracy, flow optimization goals and response timeliness. During the optimization process, flexible switching and collaborative balance between scheduling tasks are achieved through dynamic adjustment of weights. The multi-layer collaborative optimization mechanism is particularly suitable for intelligent scheduling applications of water conservancy facilities in different operation and maintenance stages and functional modes, and effectively supports comprehensive scheduling decisions under complex working conditions.

[0189] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for intelligent scheduling of water conservancy facilities based on Lévy flight optimization and BIM, characterized in that: The steps include: S1. Build a digital 3D model of the entire life cycle of the target water conservancy facility on the BIM platform; S2. Collect real-time monitoring data sets of target water conservancy facilities and map them to the digital 3D model of the entire life cycle to achieve dynamic updates of the digital 3D model data of the entire life cycle; S3. Based on the real-time, dynamically updated data on the BIM platform, extract the current operating status information and scheduling boundary conditions of the target water conservancy facilities, and establish an optimization model for the water conservancy facility scheduling problem, forming an optimization framework for the scheduling problem; S4. Initialize the control parameters of the Lévy flight optimization algorithm and generate an initial solution group based on the scheduling problem optimization model extracted from the BIM platform. At the same time, introduce an adaptive step size adjustment mechanism based on BIM semantic constraints. S5. Use the Lévy flight optimization algorithm to perform a global search for the scheduling problem, employ random jumps and adaptive adjustment mechanisms to generate multiple candidate scheduling solutions within the optimization model, and map and verify each candidate scheduling solution through the BIM platform. S6. Comprehensively evaluate the candidate scheduling plans based on the preset scheduling objectives, select the optimal scheduling plan, and map the selected optimal scheduling plan to the control system of the target water conservancy facility to generate corresponding scheduling control instructions.

2. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 1 is characterized in that: Said S1 comprises the following steps: S11. Obtain an engineering design dataset of the target water conservancy facility, including structural component numbers, geometric dimensions, material types, and spatial position parameters, and use the engineering design dataset as the structural basis for constructing a full life cycle digital 3D model to construct a structural information matrix; S12. Connect to the real-time operation monitoring system of the target water conservancy facility and collect historical and current operation status data sets, including water level monitoring values, flow records, pump station status, gate opening and closing status, and equipment start and stop time operation parameters, to construct an operation status information matrix; S13. Based on the real-time environmental perception system in the water conservancy facility operation area, collect environmental parameter data sets, including rainfall intensity, evaporation, temperature, wind speed, and soil moisture, which are external environmental information affecting the operation of water conservancy facilities, and construct an environmental parameter matrix; S14. Extract a set of key scheduling parameters, including the facility operating capacity upper limit, scheduling time window, response delay limit, and operation frequency constraint, and convert the key scheduling parameter constraint vector; S15. The structural information matrix M structure , operating status information matrix M status , environmental parameter matrix M env And the key scheduling parameter constraint vector P key Integrate into the BIM platform to build a digital 3D model of the entire life cycle of the target water conservancy facilities BIM M BIM =<M structure ,M status ,M env ,P key >。 3. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 1 is characterized in that: The S2 comprises the following steps: S21. Real-time monitoring system deployed based on water conservancy facilities collects real-time monitoring data streams Each piece of data d t Including the current water level h t , flow rate q t , device status t , valve status v t And the environmental parameter set E t , construct real-time monitoring data vector: d t =(h t ,q t ,s t ,v t ,E t ); S22. The real-time monitoring data stream D real and the structural information matrix M structure and the operating status information matrix M status Perform time tag alignment and map the acquisition time t to the full life cycle digital 3D model M in the BIM platform. BIM The corresponding running node builds the time-state mapping function: f map :d t →M BIM (t); Among them, f map Represents the real-time data mapping function, which enables each set of real-time monitoring data to be injected into the corresponding scheduling unit node; S23. Operation status information matrix M in the BIM platform status Perform dynamic updates and update the new monitoring data vector d according to the set time update cycle t Added updated operational status information matrix: in, represents the operating status information matrix before time t, Indicates updated status information; S24. Update the running status information matrix and the environmental parameter matrix M env The current window data E t Jointly inject into the BIM platform to dynamically update the digital 3D model of the entire life cycle 4. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 1 is characterized in that: The S3 includes the following steps: S31. Dynamically updated full lifecycle digital 3D model based on the BIM platform Extract the current operating status information of the target water conservancy facility and define the current operating status information vector. The current operating status information vector includes the current water level status, the current flow status, the current equipment operating status, the current valve opening and closing status, and the current environmental parameter set. S32. Based on the key scheduling parameter constraint vector P key , extract the current scheduling boundary conditions, including the scheduling capacity upper limit of the target water conservancy facilities, the allowed scheduling time window, the response delay limit and the operation frequency constraint, and construct the scheduling constraint condition set C boundary ; S33. According to the running status information vector X t+Δt and the scheduling constraint set C boundary , an optimization model for water conservancy facility scheduling is established. The optimization model defines the state space of the scheduling scheme as: Among them, S state Represents all schedulable states of the target water conservancy facilities under current constraints; S34.Based on the full life cycle digital 3D model in the BIM platform The structure and operation constraints of define the feasible solution space: in, is the i-th feasible scheduling solution, and U is the set of all possible scheduling actions; S35. Based on the actual requirements of intelligent scheduling of target water conservancy facilities, define the scheduling performance objective function of the optimization model: Among them, f level is the water level control error function, which reflects the degree to which the scheduling scheme satisfies the water level control target, f flow is the flow optimization control function, which reflects the adaptability of the scheduling scheme to flow demand, f delay is the response delay penalty function, which reflects the real-time response capability of the scheduling scheme, and w1, w2, and w3 are the weight coefficients of the scheduling target; S36. Comprehensive dispatchable state S state , feasible solution space S solution and the scheduling performance objective function f opt , and build a complete optimization framework for the intelligent scheduling of water conservancy facilities.

5. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 4 is characterized in that: The S4 comprises the following steps: S41. Full life cycle digital 3D model based on BIM platform Analyze the criticality and operational sensitivity of each dispatching node of the target water conservancy facility, and dynamically calculate the sensitivity weight coefficient η of each node based on the node's physical location, importance level and environmental impact. i ,Define the node sensitivity weight vector, and a high value of the sensitive weight coefficient indicates that the node has a greater impact on the overall safe operation and scheduling efficiency of the facility; S42. Initialize the control parameter set of the Lévy flight optimization algorithm, which includes the basic jump step factor α0, the Lévy distribution stability index β, the maximum number of iterations T max and population size N p , construct the initial control parameter set Θ LFO ; S43. Feasible solution space S according to the optimization model solution With the state space S state , generate the initial candidate solution group: Among them, each initial candidate solution It is a set of feasible scheduling strategies that meet the physical structure and operation boundary conditions of water conservancy facilities; S44. Introducing the sensitive weight coefficient η i and environmental change rateζ t Adaptive step adjustment mechanism, environmental change rate ζ t The algorithm search strategy is adjusted in real time through weighted calculation of the water level fluctuation rate, flow change trend and the impact of meteorological emergencies in real-time monitoring data. The adaptive step size is defined as: a i (t)=α0·(1+η i ·g t )。 6. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 5 is characterized in that: The S5 comprises the following steps: S51. In each round of iteration, based on the current candidate scheduling solution group X k And the corresponding adaptive step size α i (t), perform Lévy flying jump operation to generate the next generation candidate scheduling solution group X k+1 : in, It represents the i-th scheduling plan in the k-th iteration. The scheduling plan consists of a set of adjustable points and their operating states, representing a feasible scheduling path for the target water conservancy facility. represents the i-th candidate scheduling solution in the k+1-th iteration, x best represents the current global optimal scheduling solution, α i (t) is the adaptive jump step of the i-th node at time t, L β is a random variable that follows the Lévy distribution; S52. For each candidate scheduling solution generated Calculate its scheduling performance objective function S53. Calling the full life cycle digital 3D model in the BIM platform For each candidate scheduling solution Perform physical structure constraint mapping verification: Among them, the function f map (·) Determine whether the candidate solution is physically executable, including whether there are structural conflicts, path redundancies, or scheduling logic violations. A return value of True indicates that the solution structure is legal. S54. For candidate scheduling solutions whose mapping verification is False, call the structural constraint correction mechanism to perform local optimization: Among them, ∈ i is the perturbation coefficient, γ i is the node semantic weight factor, combined with the scheduling sensitivity setting of the node in the BIM model, and the key nodes are adjusted first. is the gradient direction of the objective function of the candidate scheduling solution in the local space; S55. All candidate scheduling solutions after mapping verification and local correction Recalculate its scheduling performance objective function value and compare it with the current global optimal scheduling solution x best Compare, if satisfied Then the updated global optimal scheduling solution is S56. Repeat S51 to S55 until the maximum number of iterations T is reached max , and finally obtain a set of scheduling solutions that meet the physical structure, real-time operating status and scheduling objectives of water conservancy facilities.

7. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 6 is characterized in that: The S6 comprises the following steps: S61. Based on the scheduling performance objective function For the candidate scheduling solution group X k+1 Comprehensively evaluate all legal scheduling schemes in the structure, and select the global optimal scheduling scheme based on the following strategy If all candidate solutions meet the structural constraints and scheduling boundary conditions of the water conservancy facilities, the solution with the smallest objective function value is selected: If there are multiple solutions with similar objective function values, that is, the difference is less than the set threshold, they will be further sorted according to the equipment scheduling frequency, energy consumption cost and operation complexity, and the solution with low scheduling operation frequency and simple control action set will be given priority; If operating in a high-frequency emergency environment, a real-time emergency factor θ is introduced t Dynamically modify the weight of the objective function, recalculate, and select a more responsive scheduling solution; S62. Select the optimal scheduling solution Mapped to the digital 3D model of the entire life cycle of the target water conservancy facility Each scheduling control node in the BIM model generates a control instruction set U based on the semantic label and physical connection relationship of each control node in the BIM model. ctrl ; S63. Set the control instruction set U ctrl It is sent to the control system of the target water conservancy facility to implement the selected optimal scheduling plan, start the operation status monitoring mechanism, and record the scheduling result data for subsequent model optimization and scheduling feedback adjustment.

8. The intelligent scheduling method for water conservancy facilities based on Lévy flight optimization and BIM according to claim 7 is characterized in that: The structural expression of the optimal scheduling solution is: If the water level warning value h t+Δt >h safe , and the traffic prediction value q t+Δt >q thresh , then the emergency control plan is triggered and a dispatching path with rapid flood discharge capability is selected; If the water level fluctuation rate If the energy consumption is low, the environmental parameters are stable, and the rainfall in the next 2 hours is predicted to be less than the set threshold, the scheduling plan with the lowest energy consumption is selected; If there are multiple concurrent tasks in the target area, the flow allocation priority rule is adopted to meet the downstream irrigation priority scheduling conditions; If the current device operating frequency exceeds the set maximum operating frequency F op , then the candidate solutions containing excessive operations are eliminated, and the path solutions with high structural stability are prioritized.

Images