A method and system for predicting scheduling trends based on cascade power generation in a river basin

Abstract

The present application discloses a dispatching trend prediction method and system based on cascade power generation in a river basin, which involves hydropower operation and maintenance and data processing technology, including: pre-establishing terrain models of reservoirs at all levels, and associating the terrain models in the station-level operation interface of power stations at all levels in the cascade river basin; obtaining image data in front of the reservoir dam through the dam-front monitoring equipment of the reservoirs at all levels, wherein the image data at least covers part of the reservoir water area and the reservoir boundary; identifying the water area boundary of the current reservoir based on the image data; mapping the identified water area boundary to the terrain model; in the station-level operation interface of power stations at all levels, based on the mapping state of the terrain model, and according to the dispatching data of the associated power station, the dispatching trend prediction is performed step by step. The available dispatching water is intuitively presented in a model-based manner to improve the efficiency of assisting dispatchers.

Description

Technical Field

[0001] The present application relates to the fields of hydropower operation and maintenance and data processing technology, and in particular to a method and system for predicting scheduling trends based on cascade power generation in a river basin. Background Art

[0002] With the large-scale development and access to the power grid of intermittent renewable energy sources such as photovoltaic and wind power in China, higher requirements have been placed on the peak load regulation and frequency regulation capabilities of hydropower stations. Due to the randomness, intermittency and volatility of renewable energy sources such as wind power and photovoltaic, the dispatching and operation of hydropower stations requires a more comprehensive and accurate prediction of the future operation trends than before, so as to timely predict various potential operation risks and take corresponding pre-control measures, and try to avoid adverse situations such as water level exceeding the limit and output obstruction, so as to ensure that the hydropower station group can effectively compensate and regulate various intermittent renewable energy sources at any time.

[0003] The dispatching process can be divided into two methods: determining electricity by water and determining water by electricity. Determining electricity by water means taking the available water during the dispatching period as the boundary constraint to seek the optimization solution with the maximum total power generation. Determining water by electricity means taking the active load curve or the real-time active setting value during the dispatching period as the boundary constraint to seek the optimization solution with the minimum total water consumption or energy consumption. Both methods need to consider various constraints such as water balance, upper and lower limits of reservoir water level, reservoir water level fluctuation, upper and lower limits of power station output, and upper and lower limits of power generation flow.

[0004] Regardless of the method used, the dispatching personnel are required to determine the dispatching plan for the hydropower station group and report it to the power dispatching agency for approval. The power dispatching agency will then issue the revised active load curve or real-time active setting value to the monitoring system of the hydropower station group control center, or directly to the computer monitoring system of each hydropower plant. In fact, the existing model calculation method has poor auxiliary effect on the dispatching personnel and is inefficient. Summary of the invention

[0005] The embodiments of the present application provide a method and system for predicting scheduling trends based on cascade power generation in a river basin, which is used to intuitively present the available scheduling water in a model-based manner based on the storage and scheduling plans of the reservoirs of power stations at each level, thereby improving the efficiency of assisting scheduling personnel.

[0006] The present application embodiment proposes a scheduling trend prediction method based on cascade power generation in a river basin, including:

[0007] Establishing terrain models of reservoirs at all levels in advance, and associating the terrain models with the station-level operation interfaces of power stations at all levels in the cascade basin;

[0008] Acquire image data in front of the reservoir dam through dam-front monitoring equipment of reservoirs at all levels, wherein the image data covers at least part of the reservoir water area and the reservoir boundary;

[0009] Identify the water boundary of the current reservoir based on the image data;

[0010] mapping the identified water area boundaries to the terrain model;

[0011] In the station-level operation interface of each level of power station, based on the mapping status of the terrain model and according to the dispatching data of the associated power station, the dispatching trend prediction is performed step by step.

[0012] Optionally, the image data in front of the reservoir dam is obtained through the dam front monitoring equipment of each level of reservoirs, including:

[0013] Determine the scheduling plan time of the power station, and according to each scheduling plan time, intercept the video data of a specified length from the video data collected by the monitoring equipment in front of the dam;

[0014] A plurality of image data with the highest definition are extracted from the video data as acquired image data.

[0015] Optionally, extracting a plurality of image data with the greatest definition from the video data as the acquired image data includes:

[0016] For the video data collected in an environment with good lighting conditions, multiple images with the highest definition are directly extracted as the acquired image data;

[0017] For video data collected in an environment with lighting conditions lower than a preset lighting intensity, the image area with the largest brightness is determined according to the color histogram of any video frame, and sub-images are segmented based on the image area to determine the acquired image data according to the clarity of the segmented sub-images.

[0018] Optionally, identifying the water boundary of the current reservoir based on the image data includes:

[0019] Identifying a water area according to a water area pixel interval in the acquired image data, extracting a reference boundary at a preset pixel distance outside a pixel boundary of the water area, and determining a pixel distribution between the reference boundary and the pixel boundary;

[0020] Detecting boundary data in the acquired image data using an edge detection algorithm;

[0021] dividing the boundary data into a plurality of boundary segments according to the pixel distribution;

[0022] From the pixel distributions corresponding to the multiple boundary segments, a boundary segment having a large pixel deviation from the water area pixel interval is determined as the water area boundary of the current reservoir.

[0023] Optionally, mapping the identified water area boundary to the terrain model includes:

[0024] Pre-configure a plurality of reference points in the terrain model;

[0025] Determine the position of any reference point in the image data according to the configured reference points;

[0026] According to the reference point and the boundary segment corresponding to the water area boundary, the horizontal water area data is filled in the terrain model to map the identified water area boundary to the terrain model.

[0027] Optionally, when a plurality of discontinuous boundary segments are determined from the pixel distributions corresponding to the plurality of boundary segments, the method further includes:

[0028] Selecting one of the boundary segments according to the reference point and the boundary segment corresponding to the water area boundary, and filling the horizontal water area data in the terrain model; and,

[0029] The other boundary segments are mapped to the terrain model to modify the filled water area data according to the positional relationship between the other boundary segments and the filled water area data.

[0030] Optionally, it also includes pre-configuring a time window of a preset duration, wherein the time window covers a plurality of sequential scheduling plan moments;

[0031] The boundary information of the terrain model at the previous scheduling plan time in the same period is intercepted through the time window;

[0032] According to the load information at the time of the scheduling plan in the same period and the boundary information of the terrain model, it is determined whether the boundary information of the terrain model at the current moment is within the deviation range, and if it exceeds the deviation range, the boundary information of the terrain model at the current moment is corrected.

[0033] Optionally, in the station-level operation interface of each level of power station, based on the mapping state of the terrain model and according to the dispatching data of the associated power station, the dispatching trend prediction is performed step by step, including:

[0034] Presenting the mapping status of the terrain model on the station-level operation interface of each level of power station;

[0035] Perform trend fitting according to the dispatching plan, determine the maximum water consumption for power generation based on the dispatching plan with trend fitting, and obtain the water inflow into the reservoir based on empirical data;

[0036] According to the maximum value of the water consumption for power generation and the amount of water entering the reservoir, an associated display of the scheduling trend is performed based on the presented mapping status, and the determined maximum value of the water consumption for power generation of the upper-level power station is introduced into the amount of water entering the reservoir of the lower-level power station, so as to perform scheduling trend prediction and associated display step by step, wherein the associated display is used to associate and present the increase or decrease area of ​​the water volume with the scheduling trend in a flashing manner based on the presented mapping status.

[0037] The embodiment of the present application also proposes a scheduling trend prediction system based on river basin cascade power generation, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, the steps of the scheduling trend prediction method based on river basin cascade power generation as described above are implemented.

[0038] The method of the embodiment of the present application can intuitively present the available dispatching water in a model-based manner based on the storage and dispatching plans of the reservoirs of power stations at all levels, thereby improving the efficiency of assisting dispatchers.

[0039] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0041] Figure 1 The basic process diagram of the method for predicting the scheduling trend based on cascade power generation in a river basin according to the present embodiment is shown in FIG. DETAILED DESCRIPTION

[0042] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0043] The present application embodiment proposes a scheduling trend prediction method based on cascade power generation in a river basin, such as Figure 1 As shown, the following steps are included:

[0044] In step S101, a terrain model of each level of reservoirs is pre-established, and the terrain model is associated with the station-level operation interface of each level of power station in the cascade basin. For example, GIS can be used to construct the terrain model of each level of reservoirs using elevation DEM data. After the terrain model is established, the terrain model is associated with the operation and maintenance interface of the power station, for example, based on the turbine group or dam gate, for example, by clicking on the head data to enter the terrain model interface, or the established terrain model is directly presented in the area near the turbine interface.

[0045] In step S102, image data in front of the reservoir dam is obtained through the monitoring equipment in front of the dams of various levels of reservoirs, wherein the image data at least covers part of the reservoir water area and the reservoir boundary. For example, the monitoring equipment in front of the power station dam can be directly used to obtain the image data in front of the reservoir dam.

[0046] In step S103, the water boundary of the current reservoir is identified based on the image data. In some embodiments, the water boundary of the current reservoir can be identified based on an edge detection algorithm.

[0047] In step S104, the identified water boundary is mapped to the terrain model. The embodiment of the present application establishes a terrain model and maps the terrain model based on image data. In subsequent examples, the mapping can be completed through some features of the water boundary. On the one hand, there is no need to introduce the deviation that may be caused by the head data processing of single-point detection to improve the accuracy of mapping. On the other hand, the use of monitoring equipment does not bring additional equipment costs.

[0048] In step S105, in the station-level operation interface of each level of power station, based on the mapping state of the terrain model and, according to the scheduling data of the associated power station, the scheduling trend prediction is performed step by step. In the specific example, it can be combined with the current water balance, the upper and lower limits of the reservoir water level, etc., and can also be achieved by fitting learning based on historical data combined with the mapping state of the terrain model. The step-by-step prediction referred to in the embodiment of the present application is to use the power generation and drainage of the upper power station as the storage volume of the next power station, so as to complete the available scheduling trend prediction of the current power station.

[0049] The method of the embodiment of the present application can intuitively present the available dispatching water in a model-based manner based on the storage and dispatching plans of the reservoirs of power stations at all levels, thereby improving the efficiency of assisting dispatchers.

[0050] In some embodiments, obtaining image data in front of a reservoir dam through dam-front monitoring equipment at each level of reservoirs includes:

[0051] Determine the scheduling plan time of the power station, and according to each scheduling plan time, intercept the video data of a specified length from the video data collected by the dam-front monitoring equipment. In the embodiment of the present application, the collected video data is intercepted according to each scheduling plan time, so as to realize the association between scheduling and water area data. In some examples, it can be 96 points / day corresponding scheduling plan time.

[0052] From the video data, multiple images with the highest definition are extracted as the acquired image data. For example, the definition of the image can be calculated frame by frame, so that the image for subsequent processing is selected according to the definition sorting.

[0053] In some embodiments, extracting a plurality of image data with the greatest definition from the video data as the acquired image data includes:

[0054] For video data collected under an environment with good lighting conditions, multiple images with the highest definition are directly extracted as acquired image data.

[0055] For video data collected in an environment where the illumination is lower than a preset illumination intensity, the image area with the highest brightness is determined based on the color histogram of any video frame, and sub-images are segmented based on the image area to determine the acquired image data based on the clarity of the segmented sub-images. For example, the image area with the highest brightness can be segmented to calculate only the clarity of the sub-image. In a specific application, for example, corresponding to nighttime conditions, the surveillance equipment has a higher clarity of the illuminated area of ​​the dam area collected, and the method of the present application is used to identify this area, thereby improving the accuracy of subsequent boundary recognition.

[0056] In some embodiments, identifying the water boundary of the current reservoir based on the image data includes:

[0057] In the acquired image data, a water area is identified based on a water area pixel interval, a reference boundary is extracted at a preset pixel distance outside the pixel boundary of the water area, and the pixel distribution between the reference boundary and the pixel boundary is determined. In a specific example, a water area is identified based on a water area pixel interval as a coarse boundary screening, and the size of the preset pixel distance can be set according to the pixel area of ​​the water area, so that an extended area can be selected within the water area. In the example of the present application, the pixel distribution of the selected extended area is determined, and in subsequent examples, the pixel distribution is used to correct the boundary of the image data acquired by the edge detection algorithm.

[0058] Detecting boundary data in the acquired image data using an edge detection algorithm;

[0059] dividing the boundary data into a plurality of boundary segments according to the pixel distribution;

[0060] From the pixel distributions corresponding to the multiple boundary segments, the boundary segments with large pixel deviations from the water pixel interval are determined as the water boundary of the current reservoir. In specific applications, according to the actual operating conditions, there are many interference conditions in the riverbank area of ​​the hydropower station that interfere with the detection of the water boundary using the edge detection algorithm, such as vegetation, vegetation reflections, etc. The method of the present application determines the pixel distribution of the selected extended area, thereby dividing the boundary data into multiple boundary segments, and then screening out the boundary segments with large pixel deviations from the water pixel interval, thereby eliminating the interference conditions of the environment, thereby improving the accuracy of the water boundary detected by the edge detection algorithm.

[0061] In some embodiments, mapping the identified water boundary to the terrain model comprises:

[0062] Pre-configure a plurality of reference points in the terrain model;

[0063] Determine the position of any reference point in the image data according to the configured reference points;

[0064] According to the reference point and the boundary segment corresponding to the water area boundary, the horizontal water area data is filled in the terrain model to map the identified water area boundary to the terrain model. In a specific example, the reference point can be a mapping point between the terrain model and the image data, which can be pre-set in the reservoir environment in the form of a fixed object or a fixed mark and mapped in the terrain model, so that according to the position of any reference point in the image data, the aforementioned filtered boundary segment can be used as a reference to fill the horizontal water area data in the terrain model to obtain the water area state after the terrain model is filled.

[0065] In some embodiments, when a plurality of discontinuous boundary segments are determined from the pixel distributions corresponding to the plurality of boundary segments, the method further includes:

[0066] Selecting one of the boundary segments according to the reference point and the boundary segment corresponding to the water area boundary, and filling the horizontal water area data in the terrain model; and,

[0067] Other boundary segments are mapped to the terrain model to modify the filled water area data according to the positional relationship between the other boundary segments and the filled water area data. For example, in some examples, other boundary segments are below the filled water area data, so the filling water level in the terrain model is lowered.

[0068] In some embodiments, it also includes pre-configuring a time window of a preset duration, wherein the time window covers a plurality of sequential scheduling plan moments;

[0069] The boundary information of the terrain model at the previous scheduling plan time in the same period is intercepted through the time window;

[0070] According to the load information at the time of the scheduling plan in the same period and the boundary information of the terrain model, it is determined whether the boundary information of the terrain model at the current moment is within the deviation range, and if it exceeds the deviation range, the boundary information of the terrain model at the current moment is corrected. In the embodiment of the present application, the water area information filled in the terrain model can be constrained by intercepting historical data of the same period according to the time window.

[0071] In some embodiments, in the station-level operation interface of each level of power station, based on the mapping state of the terrain model and according to the scheduling data of the associated power station, performing scheduling trend prediction step by step includes:

[0072] Presenting the mapping status of the terrain model on the station-level operation interface of each level of power station;

[0073] The trend is fitted according to the dispatching plan, the maximum water consumption for power generation is determined based on the dispatching plan with trend fitting, and the water inflow into the reservoir is obtained according to the empirical data.

[0074] According to the maximum water consumption for power generation and the amount of water entering the reservoir, the associated display of the scheduling trend is performed based on the presented mapping state, and the determined maximum water consumption for power generation of the previous power station is introduced into the inflow of the next power station, so as to perform scheduling trend prediction and associated display step by step, wherein the associated display is used to associate and present the increase and decrease areas of water volume with the scheduling trend in a flashing manner based on the presented mapping state. In a specific example, the trend prediction can simulate the operation data of each cascade reservoir according to the planned load issued at the time of the scheduling plan, combined with the latest forecast results. The specific simulation calculation can be predicted in the water balance mode, that is:

[0075]

[0076] In the formula, for Water level at the end of the period, m; for Inbound and outbound flows during the period, m³ / s; is the time period length, s; for The time period loss flow is taken as 0, m³ / s; It is the water level-reservoir capacity relationship curve, and the interpolation method adopted is linear interpolation.

[0077] When the power station / unit output is given, the steps to solve the final water level include:

[0078] A1: Set the final water level at the dispatching time;

[0079] A2: Calculate the outflow according to the water balance equation based on the inflow and the initial water level of the period;

[0080] A3: Calculate the outflow and final water level of the power station based on the output;

[0081] A4: Determine whether the calculated water level at the end of the period and the assumed value meet the accuracy requirements. If so, the calculation ends. Otherwise, the water level at the end of the period is re-assumed and the process returns to A2 until the water level at the end converges.

[0082] A5: Determine in turn whether the outflow, power station output, and final water level meet the constraint requirements. If not, calculate other power generation indicators according to the corresponding constraint values ​​and give information prompts.

[0083] The above method can be used to realize online tracking of water level, outflow and other result data, monitor the possible over-limit situation of reservoir trend based on the mapping status of the terrain model, and issue over-limit alarms to remind dispatchers to adjust plans in time, thereby greatly improving the auxiliary role for dispatchers.

[0084] The embodiment of the present application also proposes a scheduling trend prediction system based on river basin cascade power generation, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, the steps of the scheduling trend prediction method based on river basin cascade power generation as described above are implemented.

[0085] In addition, although exemplary embodiments have been described herein, the scope includes any and all embodiments based on the present disclosure with equivalent elements, modifications, omissions, combinations (e.g., various embodiments intersecting schemes), adaptations or changes. It is not limited to the examples described in this specification or during the implementation of this application, and its examples will be interpreted as non-exclusive.

[0086] The above description is intended to be illustrative rather than restrictive. For example, the above examples (or one or more of them) can be used in combination with each other. For example, those of ordinary skill in the art may use other embodiments when reading the above description.

[0087] The above embodiments are merely exemplary embodiments of the present disclosure. Those skilled in the art may make various modifications or equivalent substitutions to the present invention within the essence and protection scope of the present disclosure, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present invention.

Claims

1. A method for predicting scheduling trends based on cascade power generation in a river basin, characterized in that: include: Establishing terrain models of reservoirs at all levels in advance, and associating the terrain models with the station-level operation interfaces of power stations at all levels in the cascade basin; Acquire image data in front of the reservoir dam through dam-front monitoring equipment of reservoirs at all levels, wherein the image data covers at least part of the reservoir water area and the reservoir boundary; Identify the water boundary of the current reservoir based on the image data; mapping the identified water area boundaries to the terrain model; In the station-level operation interface of each level of power station, based on the mapping state of the terrain model and according to the dispatching data of the associated power station, the dispatching trend prediction is performed step by step; Identifying the water boundary of the current reservoir based on the image data includes: Identifying a water area according to a water area pixel interval in the acquired image data, extracting a reference boundary at a preset pixel distance outside a pixel boundary of the water area, and determining a pixel distribution between the reference boundary and the pixel boundary; Detecting boundary data in the acquired image data using an edge detection algorithm; dividing the boundary data into a plurality of boundary segments according to the pixel distribution; From the pixel distributions corresponding to the plurality of boundary segments, a boundary segment having a large pixel deviation from the water area pixel interval is determined as the water area boundary of the current reservoir; Mapping the identified water boundary to the terrain model includes: Pre-configure a plurality of reference points in the terrain model; Determine the position of any reference point in the image data according to the configured reference points; Filling horizontal water area data in the terrain model according to the reference point and the boundary segment corresponding to the water area boundary to map the identified water area boundary to the terrain model; In the case where a plurality of discontinuous boundary segments are determined from the pixel distributions corresponding to the plurality of boundary segments, the method further includes: Selecting one of the boundary segments according to the reference point and the boundary segment corresponding to the water area boundary, and filling the horizontal water area data in the terrain model; and, The other boundary segments are mapped to the terrain model to modify the filled water area data according to the positional relationship between the other boundary segments and the filled water area data.

2. The method for predicting the dispatching trend of cascade power generation in a river basin according to claim 1, characterized in that: The image data in front of the reservoir dam obtained through the dam-front monitoring equipment of reservoirs at all levels include: Determine the scheduling plan time of the power station, and according to each scheduling plan time, intercept the video data of a specified length from the video data collected by the monitoring equipment in front of the dam; A plurality of image data with the highest definition are extracted from the video data as acquired image data.

3. The method for predicting the dispatching trend of cascade power generation in a river basin according to claim 2, characterized in that: Extracting a plurality of image data with the greatest definition from the video data as the acquired image data includes: For the video data collected in an environment with good lighting conditions, multiple images with the highest definition are directly extracted as the acquired image data; For video data collected in an environment with lighting conditions lower than a preset lighting intensity, the image area with the largest brightness is determined according to the color histogram of any video frame, and sub-images are segmented based on the image area to determine the acquired image data according to the clarity of the segmented sub-images.

4. The method for predicting the dispatching trend of cascade power generation in a river basin according to claim 1, characterized in that: Also included is a time window of preconfigured duration, wherein the time window covers a plurality of sequential scheduling plan moments; The boundary information of the terrain model at the previous scheduling plan time in the same period is intercepted through the time window; According to the load information at the time of the scheduling plan in the same period and the boundary information of the terrain model, it is determined whether the boundary information of the terrain model at the current moment is within the deviation range, and if it exceeds the deviation range, the boundary information of the terrain model at the current moment is corrected.

5. The method for predicting the dispatching trend of cascade power generation in a river basin according to claim 1, characterized in that: In the station-level operation interface of each level of power station, based on the mapping state of the terrain model and according to the dispatching data of the associated power station, the dispatching trend prediction is performed step by step, including: Presenting the mapping status of the terrain model on the station-level operation interface of each level of power station; Perform trend fitting according to the dispatching plan, determine the maximum water consumption for power generation based on the dispatching plan with trend fitting, and obtain the water inflow into the reservoir based on empirical data; According to the maximum value of the water consumption for power generation and the amount of water entering the reservoir, an associated display of the scheduling trend is performed based on the presented mapping status, and the determined maximum value of the water consumption for power generation of the upper-level power station is introduced into the amount of water entering the reservoir of the lower-level power station, so as to perform scheduling trend prediction and associated display step by step, wherein the associated display is used to associate and present the increase or decrease area of ​​the water volume with the scheduling trend in a flashing manner based on the presented mapping status.

6. A dispatch trend prediction system based on cascade power generation in a river basin, characterized in that: The method comprises a processor and a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, the steps of the method for predicting the scheduling trend based on cascade power generation in a river basin as described in any one of claims 1 to 5 are implemented.

Images