Method, system and equipment for forecasting influence of flood in flood storage and detention area on power grid equipment

Abstract

The invention discloses a forecasting method, system and device for the influence of flood in a flood storage and detention area on power grid equipment, and the method comprises the steps: constructing a water level forecasting model based on a neural network method according to basic data of the flood storage and detention area and rainfall and water level historical data of a gate hydrological station, and carrying out the water level forecasting; calculating a flow process according to a gate hole flow formula, constructing a two-dimensional flooding model based on a digital elevation model and land utilization data, and simulating a two-dimensional flood flooding process under the conditions of future rainfall and upstream inflow as a flooding result; and finally, according to different time inundation ranges predicted in the flood storage and detention area and the spatial position of the power grid equipment, analyzing the affected time and inundation depth change of the power grid equipment. Compared with a traditional method for forecasting the influence of power grid equipment in rainstorm and flood, the method has the advantages that the influence degree is quantified, and the like.

Description

Technical Field

[0001] The present invention relates to the field of flood prevention and control, and in particular to a method, system and device for forecasting the impact of floods in flood storage areas on power grid equipment. Background Art

[0002] Floods occur frequently and are extremely destructive. Factors such as water scouring, soaking, and impact of floating objects in floods can easily cause power grid equipment such as transformers, switchgear, and lines to become abnormal or even damaged. Equipment damage and line interruptions can directly lead to power outages, affecting residents' lives and industrial production. Floods may also cause safety hazards such as electric shock, threatening people's lives.

[0003] The potential threat of floods to power grid equipment is indeed an urgent problem to be solved. As an important part of the flood control system, the safe operation of the power grid equipment in the flood storage area is crucial to the smooth progress of the entire flood control work. However, due to the special nature of the flood storage area, its power grid equipment is more susceptible to floods.

[0004] At present, although there are hydrological and hydraulic models with physical mechanisms for simulating and forecasting floods and waterlogging processes, these models have the following limitations: difficulty in data acquisition: the basic data required by the model, such as soil properties, river topography, etc., are difficult and costly to obtain; data missing constraints: the model is easily constrained by missing data, resulting in inaccurate forecast results; large consumption of computing resources: the model calculations are complex and consume a lot of computing resources, which is not conducive to rapid response and decision-making. Summary of the invention

[0005] Purpose of the invention: The present invention aims to provide a method for quantitatively predicting the impact of floods on power grids; another purpose of the present invention is to provide a forecast and device for the impact of floods in flood storage areas on power grid equipment.

[0006] Technical solution: The method for predicting the impact of floods in flood storage areas on power grid equipment described in the present invention comprises the following steps:

[0007] (1) Construct a water level forecasting model based on neural network method to forecast water level;

[0008] (2) Based on the digital elevation model and land use data of the flood storage area, a flood model based on the hydrodynamic model is constructed to conduct flood forecasting and obtain information on the inundation range, inundation depth, and inundation evolution process;

[0009] (3) Determine the impact of floods on power grid equipment based on the spatial location and elevation information of the power grid equipment and the flood forecast results obtained in step (2).

[0010] Furthermore, step (1) is specifically as follows:

[0011] Get historical rainfall and water level data;

[0012] Obtain a historical data set of water level and rainfall, and divide the historical data set into a training set and a validation set;

[0013] Build and train a water level forecasting model based on long short-term memory neural network;

[0014] Using the historical rainfall and water level data t hours before the present, the trained water level forecasting model is used to forecast the water level data for the next t hours.

[0015] Furthermore, step (2) is specifically as follows:

[0016] (21) Collect data elevation model and land use data and perform preprocessing;

[0017] (22) calculating the gate hole outflow rate Q according to the predicted water level obtained in step (1), and determining the gate hole outflow process according to the gate hole outflow rate Q;

[0018] (23) Establish a two-dimensional shallow water equation, construct a hydrodynamic model based on the two-dimensional shallow water equation, determine the Manning coefficient in the flood storage area based on land use data, determine the flood inundation characteristics, and construct a flood model based on the hydrodynamic model;

[0019] (24) Determine the inflow parameters according to the size parameters of the gate hole and its spatial position;

[0020] (25) The flood model performs flood forecasting based on the sluice outflow Q in step (22), the rainfall data predicted by the meteorological bureau, and the inflow parameters in step (24), and obtains information on the flooding range, flooding depth, and flooding evolution process.

[0021] Furthermore, the gate hole outflow rate Q is

[0022]

[0023] Among them, σ is the flooding coefficient, μ is the flow coefficient, e is the gate opening height, n is the number of gate holes, b is the net width of the gate hole, g is the gravitational acceleration, and H is the head above the weir, that is, the elevation difference between the water level above the gate and the water level below the gate.

[0024] Furthermore, the water head H above the weir is determined by the water level predicted in step (1) and the water level difference in the flood storage area.

[0025] Furthermore, in step (22), the two-dimensional shallow water equation is

[0026]

[0027] Where h is the water depth, t is the time, u and v are the flow velocities in the x and y directions respectively, Sce is the source or sink of the flow diagram, Z is the free surface elevation, and v e is the effective viscosity coefficient, g is the gravitational acceleration, F x is the friction term in the x direction, F y is the friction term in the y direction.

[0028] Furthermore, step (3) is as follows:

[0029] Obtain the spatial location information of power grid equipment and its elevation information in the elevation model coordinate system;

[0030] According to the spatial position of the power grid facility and the flood forecast result obtained in step (2), the corresponding grid position of the flood result is determined, and the time variation sequence of the water depth at the corresponding position is obtained;

[0031] According to the elevation attributes of the power grid facilities themselves, the affected time and submerged water depth of the power grid facilities are determined.

[0032] The forecasting system for the impact of floods in flood storage areas on power grid equipment of the present invention comprises:

[0033] The water level forecast model building module is used to build a water level forecast model based on the neural network method to forecast water level;

[0034] The flood model construction module is used to construct a flood model based on a hydrodynamic model according to the digital elevation model and land use data of the flood storage area, conduct flood forecasting, and obtain information on the inundation range, inundation depth, and inundation evolution process;

[0035] The prediction module of the impact of power grid equipment is used to determine the impact of floods on power grid equipment based on the spatial location and elevation information of the power grid equipment and the flood forecast results obtained by the flood model construction module.

[0036] The computer device of the present invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the steps of the above method are implemented when the processor executes the computer program.

[0037] The computer-readable storage medium of the present invention stores a computer program thereon, and the computer program implements the steps of the above method when executed by a processor.

[0038] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: 1. The present invention is based on a method for forecasting and analyzing flood impacts on power grids in flood storage and detention areas based on a combination of a neural network and a physical mechanism model. The neural network model is used to replace the physical model with a large demand for basic data, which not only reduces the cost of modeling, but also improves the calculation efficiency of flood forecasting. The forecast results of flood impacts can be quickly obtained in combination with the attributes of power grid facilities, which has a positive effect on ensuring the safety of electricity use in flood storage and detention areas; 2. The present invention adopts a neural network method to replace part of the simulation model based on the physical mechanism, which reduces the impact of the lack of basic data on modeling, and improves the efficiency of simulation and forecasting calculations; 3. The present invention quantifies the impact of floods on power grid facilities, has a more accurate guiding role in flood prevention preparation measures for power grid facilities in flood storage and detention areas, and improves the guarantee of safe operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the process of the present invention;

[0040] Figure 2 This is a schematic diagram of the structure of a long short-term memory neural network;

[0041] Figure 3 Flowchart for flood simulation forecast;

[0042] Figure 4 The flowchart for the analysis of the impact on power grid equipment is as follows: DETAILED DESCRIPTION

[0043] The present invention will be further described below in conjunction with the accompanying drawings.

[0044] The method for forecasting the impact of floods in a flood storage area on power grid equipment of the present invention comprises the following steps:

[0045] (1) Based on historical rainfall and water level data, a historical data set and a future prediction data set are constructed, a long-term and short-term memory neural network model is trained, and a water level forecasting model based on the neural network method is constructed.

[0046] (11) Obtain historical rainfall and water level data.

[0047] Specifically, due to the gap between data from different data sources, the time interval of historical rainfall and water level data is processed to 1 hour according to the forecast requirements. In addition, it is also necessary to perform operations on missing or extremely abnormal data.

[0048] (12) Construct historical data sets of water level and rainfall and corresponding prediction data sets.

[0049] Specifically, the historical data is divided into a 72-hour historical data matrix and a corresponding relatively future 72-hour data matrix according to the forecast time requirements, and the historical data set and forecast data are established.

[0050] (13) Based on the constructed data set, a long short-term memory neural network model is constructed and trained.

[0051] By reading the output of the previous time step and the input of the current time step of the LSTM neural network, and then using them as the input of the forget gate, input gate and output gate of the LSTM neural network.

[0052] Specifically, the output of the LSTM neural network at the previous time step refers to the water level predicted at the previous time step, and the input data refers to a two-dimensional array consisting of historical precipitation data and historical water levels.

[0053] The forget gate of the LSTM neural network determines what information is discarded from the cell state.

[0054] Specifically, the cell state is the key to the long short-term memory neural network's suitability for long-range dependencies, as it stores historical information. The forget gate first maps the input into a vector with each dimension value between 0 and 1; then the vector is multiplied by the cell state, and the part multiplied by the value close to 0 in the vector is discarded.

[0055] The input gate of the LSTM neural network determines what information is stored from the cell state.

[0056] Specifically, the input gate limits the input value to between -1 and 1 through a function, which helps to regulate the value flowing through the network, and at the same time maps the input into a vector with each dimension value between 0 and 1; then multiplies the vector with the output of the function to obtain the information that needs to be updated; finally, the information is added to the forgotten cell state to obtain the new cell state.

[0057] The output result type is determined by the output gate of the long short-term memory neural network.

[0058] Specifically, the output gate inputs the cell state into the function; then it maps the input into a vector with each dimension value between 0 and 1; finally, the vector is multiplied by the output of the function to determine the value to be output, and at this time it will continue to be input into the next time step for further calculation.

[0059] (14) Based on the trained model and historical data near the forecast time, the water level process in the next 72 hours is predicted.

[0060] Specifically, for a specific water level station, the trained long short-term memory neural network model parameters are fixed, the rainfall and water level data 72 hours before the start of the forecast are input, and the water level process data for the next 72 hours are output.

[0061] (2) Based on the inflow process data calculated from the forecast water level data, a hydrodynamic model is constructed based on the digital elevation model, land use data, and other data to perform simulation forecasts and obtain flood forecast results.

[0062] Obtain data elevation model and land use data.

[0063] Specifically, after obtaining the digital elevation model and land use data of the simulated flood storage area, it is necessary to unify the projection coordinate system and perform preprocessing. For example, for the digital elevation model, the elevation of the dam location is artificially raised according to the scope of the flood storage area.

[0064] The outflow rate Q of the gate hole is

[0065]

[0066] Among them, σ is the flooding coefficient, which takes the value of 1 during free outflow, e is the gate opening height, n is the number of gate holes, b is the net width of the gate hole, g is the gravitational acceleration, and H is the head on the weir, that is, the water level difference inside and outside the gate hole, which is determined by the predicted water level and the water level difference in the flood storage area.

[0067] Specifically, in the first hour after the gate is opened, the ground elevation or initial water level in the flood storage area is used to determine. After the simulation starts, the water level height of the grid point inside the gate is used for the height difference calculation.

[0068] The two-dimensional shallow water equation is

[0069]

[0070] Where h is the water depth, t is the time, u and v are the flow velocities in the x and y directions respectively, Sce is the source or sink of the flow diagram, Z is the free surface elevation, and v e is the effective viscosity coefficient, g is the gravitational acceleration, F x is the friction term in the x direction, F y is the friction term in the y direction.

[0071] The hydrodynamic model is constructed based on the shallow water equation, and the Manning coefficient in the flood storage area is determined based on the land use data and set as the characteristic value. Specifically, the distribution of the Manning coefficient in the simulation area is determined based on the spatial distribution of land use and the corresponding relationship between land use and the Manning coefficient.

[0072] The inflow position and parameters are determined according to the spatial position and characteristics of the gate hole. Specifically, the position of the inflow line is determined according to the spatial position of the gate hole. The parameters of the inflow line are determined according to the width, height, etc. of the gate hole and the terrain characteristics of the location.

[0073] According to the two-dimensional flooding model and inflow process in the region, flood parameter simulation is performed on the flood inundation characteristic data to obtain flood simulation forecast data. The obtained flow forecast data and forecast rainfall data are input into the two-dimensional flood inundation model to simulate and forecast the two-dimensional inundation results of the future flood storage area. The output of the result is a two-dimensional grid data per hour in the next 72 hours.

[0074] (3) Based on the spatial location of power grid facilities, their own attributes and the superposition analysis of flood forecast results, the impact forecast of power grid facilities is obtained.

[0075] Acquire data such as the spatial position and elevation of power grid equipment; specifically, acquire the spatial position information of the target power grid equipment and the elevation information of the equipment in the corresponding digital elevation model coordinate system.

[0076] According to the spatial position of the power grid facility, the corresponding grid position of the flooding result is determined, and the time variation sequence of the water depth at the corresponding position is obtained; specifically, according to the spatial position information of the power grid equipment, the corresponding grid is determined, and the flooding water depth of the corresponding grid is read in chronological order to form the time series data of the flooding water depth at the power grid facility location.

[0077] According to the elevation attributes of the power grid facilities themselves, the affected time and flooded water depth of the power grid facilities are determined; specifically, based on the actual elevation of the power grid facilities and the flooded water depth, the time when the water depth exceeds the facility height is predicted, and the change in flooded depth is determined.

[0078] The forecasting system for the impact of floods in flood storage areas on power grid equipment of the present invention comprises:

[0079] The water level forecast model building module is used to build a water level forecast model based on the neural network method to forecast water level;

[0080] The flood model construction module is used to construct a flood model based on a hydrodynamic model according to the digital elevation model and land use data of the flood storage area, conduct flood forecasting, and obtain information on the inundation range, inundation depth, and inundation evolution process;

[0081] The prediction module of the impact of power grid equipment is used to determine the impact of floods on power grid equipment based on the spatial location and elevation information of the power grid equipment and the flood forecast results obtained by the flood model construction module.

[0082] The computer device of the present invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the steps of the above method are implemented when the processor executes the computer program.

[0083] The computer-readable storage medium of the present invention stores a computer program thereon, and the computer program implements the steps of the above method when executed by a processor.

Claims

1. A method for predicting the impact of floods in flood storage areas on power grid equipment, characterized in that: The following steps are involved: (1) Construct a water level forecasting model based on neural network method to forecast water level; (2) Based on the digital elevation model and land use data of the flood storage area, a flood model based on the hydrodynamic model is constructed to conduct flood forecasting and obtain information on the inundation range, inundation depth, and inundation evolution process; (3) Determine the impact of floods on power grid equipment based on the spatial location and elevation information of the power grid equipment and the flood forecast results obtained in step (2).

2. The method for predicting the impact of floods in flood storage areas on power grid equipment according to claim 1, characterized in that: Step (1) is as follows: Get historical rainfall and water level data; Obtain a historical data set of water level and rainfall, and divide the historical data set into a training set and a validation set; Build and train a water level forecasting model based on long short-term memory neural network; Using the historical rainfall and water level data t hours before the present, the trained water level forecasting model is used to forecast the water level data for the next t hours.

3. The method for predicting the impact of floods in flood storage areas on power grid equipment according to claim 1, characterized in that: Step (2) is as follows: (21) Collect data elevation model and land use data and perform preprocessing; (22) calculating the gate hole outflow rate Q according to the predicted water level obtained in step (1), and determining the gate hole outflow process according to the gate hole outflow rate Q; (23) Establish a two-dimensional shallow water equation, construct a hydrodynamic model based on the two-dimensional shallow water equation, determine the Manning coefficient in the flood storage area based on land use data, determine the flood inundation characteristics, and construct a flood model based on the hydrodynamic model; (24) Determine the inflow parameters according to the size parameters of the gate hole and its spatial position; (25) The flood model performs flood forecasting based on the sluice outflow Q in step (22), the rainfall data predicted by the meteorological bureau, and the inflow parameters in step (24), and obtains information on the flooding range, flooding depth, and flooding evolution process.

4. The method for predicting the impact of floods in flood storage areas on power grid equipment according to claim 3 is characterized in that: The gate hole outflow rate Q is Among them, σ is the flooding coefficient, μ is the flow coefficient, e is the gate opening height, n is the number of gate holes, b is the net width of the gate hole, g is the gravitational acceleration, and H is the head above the weir, that is, the elevation difference between the water level above the gate and the water level below the gate.

5. The method for predicting the impact of floods in flood storage areas on power grid equipment according to claim 4, characterized in that: The water head H above the weir is determined by the water level predicted in step (1) and the water level difference in the flood storage area.

6. The method for predicting the impact of floods in flood storage areas on power grid equipment according to claim 3, characterized in that: In step (22), the two-dimensional shallow water equation is Where h is the water depth, t is the time, u and v are the flow velocities in the x and y directions respectively, Sce is the source or sink of the flow diagram, Z is the free surface elevation, and v e is the effective viscosity coefficient, g is the acceleration due to gravity, F x is the friction term in the x direction, F y is the friction term in the y direction.

7. The method for predicting the impact of floods in flood storage areas on power grid equipment according to claim 1, characterized in that: Step (3) is as follows: Obtain the spatial location information of power grid equipment and its elevation information in the elevation model coordinate system; According to the spatial position of the power grid facility and the flood forecast result obtained in step (2), the corresponding grid position of the flood result is determined, and the time variation sequence of the water depth at the corresponding position is obtained; According to the elevation attributes of the power grid facilities themselves, the affected time and submerged water depth of the power grid facilities are determined.

8. A forecasting system for the impact of floods in flood storage areas on power grid equipment, characterized in that: include The water level forecast model building module is used to build a water level forecast model based on the neural network method to forecast water level; The flood model construction module is used to construct a flood model based on a hydrodynamic model according to the digital elevation model and land use data of the flood storage area, conduct flood forecasting, and obtain information on the inundation range, inundation depth, and inundation evolution process; The prediction module of the impact of power grid equipment is used to determine the impact of floods on power grid equipment based on the spatial location of the power grid equipment, elevation information and the flood forecast results obtained by the flood model construction module.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

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

Images