A method and system for early warning of downstream water level in power plants based on sudden changes in power output

Abstract

This invention provides a method for early warning of downstream water level in a power plant based on sudden changes in power output, comprising: acquiring real-time power output data of the power plant and real-time water level data at an early warning point; triggering an early warning and issuing an early warning message in response to the power output change data of the power plant meeting a first preset condition; wherein, the early warning message includes a water level change delay time and / or a water level change amplitude; the determination of the water level change delay time includes: based on the real-time water level data, determining a target time point at which the real-time water level data meets a second preset condition according to an early warning model; and using the difference between the target time point and the early warning time point at which the early warning message is issued as the water level change delay time.

Description

Technical Field

[0001] This invention relates to the field of downstream water level early warning for power plants, and specifically to a method, system, device, and storage medium for early warning of downstream water level for power plants based on sudden changes in power output. Background Technology

[0002] Downstream water levels at the dam site of a hydroelectric power station typically need to remain stable to ensure navigation and daily life downstream. During the operation of the power station, the power output is the main factor affecting downstream water levels, and changes in downstream water levels begin in the waters near the dam and gradually spread to further downstream areas.

[0003] Under current conditions, planned power output adjustments can be announced in advance by relevant management departments to inform downstream units and riverside residents, allowing them to prepare contingency plans. If a sudden event occurs during power plant operation, causing a sudden drop in output (usually due to unit shutdown leading to a sudden decrease in output), resulting in a sudden change in the downstream water level, the power plant's operation and management unit will promptly notify the corresponding downstream management unit of the water level change warning. However, because this information transmission involves human intervention, timeliness cannot be guaranteed.

[0004] Meanwhile, the distances between the downstream locations that require early warning and the power station are different, and the delay time for water level changes is also different, resulting in different preparation times for emergency measures to be taken at different downstream locations.

[0005] The current early warning information does not include the crucial data of the expected time of water level changes, which may lead to untimely and inadequate emergency preparedness.

[0006] Therefore, there is a need for a method, system, device, and storage medium for downstream water level early warning of power plants based on sudden output changes. This method uses changes in power plant output as the trigger signal for water level change early warning information, and takes advantage of the principle that the transmission speed of water level changes is much slower than the speed of information propagation. The early warning time is approximately the lag time from the change in power output to the water level change at the warning location. Summary of the Invention

[0007] The purpose of this invention is to provide a method for early warning of downstream water levels in power plants based on sudden changes in power output, in order to solve the technical problems existing in the background art.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A method for early warning of downstream water level in a power plant based on sudden changes in power output includes:

[0010] Acquire real-time power output data of the power plant and real-time water level data of early warning points;

[0011] When the power plant output change data meets the first preset condition, an early warning is triggered and an early warning message is issued.

[0012] The early warning information includes the water level change delay time and / or the water level change amplitude; the determination of the water level change delay time includes:

[0013] Based on the real-time water level data, and according to the early warning model, the target time point at which the real-time water level data meets the second preset condition is determined.

[0014] The difference between the target time point and the warning time point when the warning information was issued is used as the water level change delay time.

[0015] In some embodiments, the generation of the early warning model includes:

[0016] Obtain basic relationship data;

[0017] Based on the aforementioned basic relationship data and a preset fitting method, a fitting function is generated as the initial early warning model.

[0018] Based on the measured data, the initial early warning model is iteratively optimized to obtain the early warning model.

[0019] In some embodiments, the basic relationship data includes:

[0020] Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the first change value, and the start time of the sudden change in output is recorded until the first time when the water level change at the warning location reaches the preset change value.

[0021] Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the second change value, and the start time of the sudden change in output is recorded until the second time when the water level change at the warning location reaches the preset change value.

[0022] Under the condition that the power plant output is stable at the second preset value, within the first preset time period, the power plant output is reduced by the first change value, and the time of the sudden change in output is recorded until the third time when the water level change at the warning location reaches the preset change value.

[0023] In some embodiments, the measured data includes:

[0024] The actual measured power plant output data before the warning is triggered, the actual power plant output change data, and the actual water level change delay time, wherein the actual water level change delay time is the difference between the actual time point when the real-time water level data meets the second preset condition and the warning time point when the warning information is issued.

[0025] In some embodiments, the triggering of the warning includes:

[0026] The early warning information includes the estimated arrival time T of the water level change and the amount of water level change H, both of which are updated regularly.

[0027] Among them, the estimated time of water level change is displayed one minute less than the previous refresh time each time it is refreshed, until the displayed value becomes 0;

[0028] The water level change H is refreshed every 2 minutes. Each time it is refreshed, the value displayed is the difference between the water level corresponding to the maximum output value before the power output change and the water level corresponding to the current power output value of the power station.

[0029] The timer starts from the first release of the warning information, and real-time water level data of the warning point is acquired and stored every second for a storage period of t0;

[0030] Compare the water level value Lt0 at the warning location before time t0 with the current real-time water level value. If the difference exceeds the limit h0, stop timing. The recorded duration is the actual water level change delay time under this sudden change in output.

[0031] In some embodiments, the early warning model is a machine learning model;

[0032] The step of determining the target time point where the real-time water level data meets the second preset condition based on the early warning model includes:

[0033] Based on the real-time water level data and the early warning water level data, the target time point is determined through the early warning model.

[0034] In some embodiments, the first preset condition includes: the power output change data of the power plant exceeds 400MW within 10 seconds.

[0035] Meanwhile, this invention also discloses a power station downstream water level early warning system based on sudden output changes, comprising:

[0036] The acquisition module is used to acquire real-time power output data of the power plant and real-time water level data of the early warning points;

[0037] The early warning module is used to trigger an early warning and issue an early warning message when the power output change data of the power plant meets the first preset condition;

[0038] The early warning information includes the water level change delay time and / or the water level change amplitude; the determination of the water level change delay time includes:

[0039] Based on the real-time water level data, and according to the early warning model, the target time point at which the real-time water level data meets the second preset condition is determined.

[0040] The difference between the target time point and the warning time point when the warning information was issued is used as the water level change delay time.

[0041] Meanwhile, the present invention also discloses a power station downstream water level early warning device based on power output change, the device including at least one processor and at least one memory;

[0042] The at least one memory is used to store computer instructions;

[0043] The at least one processor is used to execute at least some of the instructions in the computer instructions to implement the above-described method for early warning of downstream water level of power plants based on sudden changes in output.

[0044] In addition, the present invention also discloses a computer-readable storage medium that stores computer instructions that, when executed by a processor, implement the above-described method.

[0045] Beneficial effects

[0046] The significant advantages of this invention compared to existing technologies are:

[0047] The technical solution of this invention uses power plant output changes as the early warning trigger condition. Since output changes determine water level changes, and the output value has the characteristic of real-time acquisition and transmission, changes in the downstream water level can be monitored before they occur, ensuring the timeliness and accuracy of early warning triggering. Specifically, this technical solution can obtain the most important early warning data, such as the water level change delay time T and the water level change amplitude H, using only the power plant output value as a single data point. Therefore, the technical solution of this invention requires fewer data sampling devices and has strong adaptability. It is applicable to any downstream location affected by the power plant, without needing to consider geographical conditions such as the distance between the location and the power plant or the river flow direction.

[0048] Meanwhile, as the number of warnings increases, the warning model can be continuously optimized, and the warning information will become increasingly accurate. Once the model matures and stabilizes, the water level gauges at the warning locations can be omitted, further simplifying the structure of the warning system and saving investment. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the downstream water level early warning system for power plants based on sudden power output changes, which is involved in this embodiment.

[0050] Figure 2 This is a schematic diagram illustrating the application scenario of the downstream water level early warning system for power plants involved in this embodiment;

[0051] Figure 3 This is a flowchart illustrating the downstream water level early warning method for power plants based on sudden power output changes involved in this embodiment;

[0052] Figure 4 This is a flowchart illustrating a method for early warning of downstream water levels in a power station, as described in one embodiment.

[0053] Figure 5 This is a schematic diagram of the early warning model involved in this embodiment;

[0054] Figure 6 This is a flowchart illustrating a single early warning process involved in this embodiment. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] Conversely, this application covers any substitutions, modifications, equivalent methods, and schemes made within the spirit and scope of this application as defined in the claims. Furthermore, to provide the public with a better understanding of this application, certain specific details are described in detail below. However, this application can be fully understood by those skilled in the art even without these detailed descriptions.

[0057] The following will combine Figure 1-3 This application provides a detailed description of a method for early warning of downstream water levels in a power plant based on sudden changes in power output, as illustrated in the embodiments of this application. It is worth noting that the following embodiments are merely illustrative of this application and do not constitute a limitation thereof.

[0058] like Figure 1 As shown, the present invention also discloses a power plant downstream water level early warning system based on sudden output changes, comprising:

[0059] The acquisition module 110 is used to acquire real-time power output data of the power station and real-time water level data of the early warning point;

[0060] The early warning module 120 is used to trigger an early warning and issue an early warning message in response to the power plant output change data meeting a first preset condition; in some embodiments, the first preset condition includes: the power plant output change data exceeds 400MW within 10 seconds.

[0061] The early warning information includes the water level change delay time and / or the water level change amplitude; the determination of the water level change delay time includes:

[0062] Based on the real-time water level data, a target time point is determined according to the early warning model to ensure that the real-time water level data meets the second preset condition; in some embodiments, the second preset condition may include the real-time water level reaching the early warning water level.

[0063] The difference between the target time point and the warning time point when the warning information was issued is used as the water level change delay time.

[0064] In some embodiments, the generation of the early warning model includes:

[0065] Obtain basic relationship data;

[0066] Based on the aforementioned basic relationship data and a preset fitting method, a fitting function is generated as the initial early warning model.

[0067] Based on the measured data, the initial early warning model is iteratively optimized to obtain the early warning model.

[0068] In some embodiments, the basic relationship data includes:

[0069] Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the first change value, and the start time of the sudden change in output is recorded until the first time when the water level change at the warning location reaches the preset change value.

[0070] Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the second change value, and the start time of the sudden change in output is recorded until the second time when the water level change at the warning location reaches the preset change value.

[0071] Under the condition that the power plant output is stable at the second preset value, within the first preset time period, the power plant output is reduced by the first change value, and the time of the sudden change in output is recorded until the third time when the water level change at the warning location reaches the preset change value.

[0072] In some embodiments, the measured data includes:

[0073] The actual measured power plant output data before the warning is triggered, the actual power plant output change data, and the actual water level change delay time, wherein the actual water level change delay time is the difference between the actual time point when the real-time water level data meets the second preset condition and the warning time point when the warning information is issued.

[0074] In some embodiments, the triggering of the warning includes:

[0075] The early warning information includes the estimated arrival time T of the water level change and the amount of water level change H, both of which are updated regularly.

[0076] Among them, the estimated time of water level change is displayed one minute less than the previous refresh time each time it is refreshed, until the displayed value becomes 0;

[0077] The water level change H is refreshed every 2 minutes. Each time it is refreshed, the value displayed is the difference between the water level corresponding to the maximum output value before the power output change and the water level corresponding to the current power output value of the power station.

[0078] The timer starts from the first release of the warning information, and real-time water level data of the warning point is acquired and stored every second for a storage period of t0;

[0079] Compare the water level value Lt0 at the warning location before time t0 with the current real-time water level value. If the difference exceeds the limit h0, stop timing. The recorded duration is the actual water level change delay time under this sudden change in output.

[0080] In some embodiments, the early warning model is a machine learning model;

[0081] The step of determining the target time point where the real-time water level data meets the second preset condition based on the early warning model includes:

[0082] Based on the real-time water level data and the early warning water level data, the target time point is determined using an early warning model. For further explanation of the early warning model, please refer to [link / reference needed]. Figure 5 Related content.

[0083] like Figure 2 The diagram illustrates an application scenario for a power plant downstream water level early warning system 100 based on sudden output changes. In some embodiments, the power plant downstream water level early warning system 100 based on sudden output changes can be deployed in, for example... Figure 2 The warning center shown is 20.

[0084] The early warning center 20 can collect power plant output data transmitted from the power plant output monitoring station 10. A water level monitoring station 40 can be set up at the early warning location 30, and the water level data collected by the water level monitoring station 40 is also transmitted to the early warning center 20 in real time. The early warning center 20 analyzes and processes the received data. If a change in power output triggers an early warning condition, the early warning center 20 will immediately transmit the early warning information to the early warning notification terminal 50 set up at the early warning location 30, thus achieving early warning.

[0085] like Figure 3 As shown, a method for early warning of downstream water level in a power station based on sudden changes in power output includes:

[0086] Step 310: Obtain real-time power output data of the power station and real-time water level data of the early warning point.

[0087] In some embodiments, according to Figure 2 The early warning center 20, built on the architecture, acquires power plant output and early warning point water level data with a unified time coordinate.

[0088] Step 320: In response to the power plant output change data meeting the first preset condition, an early warning is triggered and an early warning message is issued.

[0089] In some embodiments, the first preset condition includes: the power output change data of the power plant exceeds 400MW within 10 seconds.

[0090] The early warning information includes the water level change delay time and / or the water level change amplitude; the determination of the water level change delay time includes:

[0091] Step 1: Based on the real-time water level data, determine the target time point when the real-time water level data meets the second preset condition according to the early warning model;

[0092] In some embodiments, the second preset condition may include the real-time water level reaching the warning water level.

[0093] Step 2: The difference between the target time point and the warning time point when the warning information was issued is taken as the water level change delay time.

[0094] In some embodiments, the generation of the early warning model includes:

[0095] S1. Obtain basic relationship data.

[0096] In some embodiments, the basic relationship data includes:

[0097] Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the first change value, and the start time of the sudden change in output is recorded until the first time when the water level change at the warning location reaches the preset change value.

[0098] Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the second change value, and the start time of the sudden change in output is recorded until the second time when the water level change at the warning location reaches the preset change value.

[0099] Under the condition that the power plant output is stable at the second preset value, within the first preset time period, the power plant output is reduced by the first change value, and the time of the sudden change in output is recorded until the third time when the water level change at the warning location reaches the preset change value.

[0100] For example, the following three experiments can be conducted to obtain basic relationship data on the relationship between output Q and the delay time T of water level change at the warning location:

[0101] 1) Under the initial condition that the power plant output is stable at Q1, conduct output adjustment test 1, and reduce the power plant output q1 within the time period t1. Record the time from the start of the output change until the water level change at the warning location reaches s as T1.

[0102] 2) Under the initial condition that the power plant output is stable at Q1, conduct output adjustment test 2, reducing the power plant output q2 within the time period t1. Record the time from the start of the output change until the water level change at the warning location reaches s as T2.

[0103] 3) Under the initial condition that the power plant output is stable at Q2, conduct output adjustment test 3, and reduce the power plant output q1 within the time period t1. Record the time from the start of the output change until the water level change at the warning location reaches s as T3.

[0104] Here, we define the power plant output as Q, the water level at the warning location as L, and the change in output Q under a sudden change in operating conditions as q. The criterion for determining if the water level at the warning location has changed is the maximum tolerable water level change per unit time, i.e., the maximum amplitude. A water level change exceeding the maximum amplitude will adversely affect production and daily life at the warning location. The maximum tolerable amplitude may vary for different warning locations. Assume that the acceptable maximum water level amplitude s at a certain warning location is the water level L change h0 within time t0. The water level change delay time T is the time taken from the start of the power plant output change to the downstream water level change at the warning location reaching s. Both the output value Q before the change and the output change q affect the water level change delay time T; therefore, the relationship between the three can be represented by the function T = f(Q, q).

[0105] S2. Based on the aforementioned basic relationship data and a preset fitting method, generate a fitting function as the initial early warning model.

[0106] For example, the above power output change test can be combined with the power plant unit trip test. After the test is completed, the above three sets of data can be used to fit the function T = f(Q,q) to generate the original early warning model. The fitting method can be the common least squares method.

[0107] S3. Based on the measured data, iteratively optimize the initial early warning model to obtain the early warning model.

[0108] In some embodiments, the measured data includes:

[0109] The actual measured power plant output data before the warning is triggered, the actual power plant output change data, and the actual water level change delay time, wherein the actual water level change delay time is the difference between the actual time point when the real-time water level data meets the second preset condition and the warning time point when the warning information is issued.

[0110] For example, the power plant output value Q, the sudden output change value q, and the actual water level change delay time Ts before the warning was triggered can be used as a new set of data to fit and optimize the warning model. The real data obtained after each warning is triggered can be used to optimize the original warning model.

[0111] In some embodiments, the triggering of the warning includes:

[0112] The early warning information includes the estimated arrival time T of the water level change and the amount of water level change H, both of which are updated regularly.

[0113] Among them, the estimated time of water level change is displayed one minute less than the previous refresh time each time it is refreshed, until the displayed value becomes 0;

[0114] The water level change H is refreshed every 2 minutes. Each time it is refreshed, the value displayed is the difference between the water level corresponding to the maximum output value before the power output change and the water level corresponding to the current power output value of the power station.

[0115] The timer starts from the first release of the warning information, and real-time water level data of the warning point is acquired and stored every second for a storage period of t0;

[0116] Compare the water level value Lt0 at the warning location before time t0 with the current real-time water level value. If the difference exceeds the limit h0, stop timing. The recorded duration is the actual water level change delay time under this sudden change in output.

[0117] In some embodiments, if the warning is triggered before the power output mutation test is fully completed and the warning model is generated, the historical delay time data T that has been acquired can be directly used for the warning.

[0118] For further explanation on determining the target time point based on the early warning model, please refer to [link / reference]. Figure 5 Related content.

[0119] like Figure 4 The diagram shows a flowchart of a downstream water level early warning method for a power station in one embodiment. Figure 4 As shown, the downstream water level early warning method for power stations may include:

[0120] The first step is to obtain real-time power output data of the power station and water level data at the warning location.

[0121] The second step is to obtain basic data on the relationship between typical output changes and water level change delays through output mutation tests.

[0122] The third step is to generate the original early warning model and set the early warning conditions.

[0123] In some embodiments, the warning condition may include a power output change exceeding a preset value q0 within a unit time t1. The warning condition is typically the tripping condition of the power plant's generator units, for example, a power output change exceeding 400MW within 10 seconds.

[0124] It should be noted that the early warning conditions can also be adapted to the actual situation. For example, in addition to power plant output data, the early warning center can also collect information on changes in the opening of floodgates, which can be used as early warning conditions for downstream water level changes during flood discharge. By changing the early warning trigger conditions, the early warning center can monitor other non-constant operating conditions besides sudden output changes (unit shutdown) and issue alerts before water level changes occur at the warning location.

[0125] The fourth step is to trigger an early warning by predicting the water level change delay based on the original early warning model.

[0126] In some embodiments, if the power output change reaches the early warning trigger condition, the predicted delay time T is obtained using the early warning model and sent to the early warning notification terminal. If the power output change test is not fully completed, the obtained delay time data T is directly used for early warning.

[0127] The fifth step is to obtain the actual water level change delay for this warning.

[0128] For example, it can monitor the actual water level changes at the warning location and record the delay time Ts from the issuance of the warning information to the water level change at the warning location reaching s.

[0129] Step 6: Iterative optimization of the early warning model.

[0130] For example, the power plant output value Q, the sudden output change value q, and the actual water level change delay time Ts before the warning was triggered can be used as a new set of data to fit and optimize the warning model. The real data obtained after each warning is triggered can be used to optimize the original warning model.

[0131] like Figure 5 The diagram shown illustrates the publicly disclosed early warning model implemented in this paper. Figure 5 As shown, in some embodiments, the warning model 520 can be a machine learning model, such as a model built from neural networks (NNs) or other intelligent learning networks.

[0132] In some embodiments, the early warning model can be trained based on a large amount of first training data 540 (including first training samples with first labels). The first training samples can be historically acquired real-time water level data 510 before the early warning. The first label can be the time point corresponding to when the water level reaches the warning value after the early warning for the first training sample.

[0133] In some embodiments, the initial warning model 550 can be trained to obtain the warning model 520. For example, the first training sample is input into the initial warning model 550 to obtain the target time point 530 corresponding to the first training sample. During training, the warning model can construct a loss function based on the labels and output results. At the same time, the parameters of the initial warning model 550 can be updated until preset conditions are met, and training is completed. The preset conditions may include one or more of the following: the loss function is less than a threshold, convergence, or the training period reaches a threshold.

[0134] like Figure 6 The diagram shown illustrates the process of a single early warning in this embodiment. Figure 6 As shown, the implementation process for a single early warning is as follows:

[0135] 1) Acquire real-time power output data Q of the power station once per second and store it for a storage time of t1.

[0136] 2) Sort the power output of the power station obtained within the time period t1 by size, and sort the output once every 1 second.

[0137] 3) Compare the first data Qx (maximum value) after sorting with the last data Qi (minimum value). The difference between the two is the output change value q. The comparison frequency is once per second.

[0138] 4) Determine if q is greater than the warning trigger setting value q0. If it is greater than q0, then trigger the warning; if it is less than q0, then do not trigger the warning. The judgment frequency is usually once per second.

[0139] 5) After the warning is triggered, the maximum output value Qx when the warning is triggered and the output change value q (i.e. Qx-Qi) when the warning is triggered are used as the input conditions of the warning model to obtain the predicted water level change delay time T at the warning location.

[0140] 6) Based on the power station's "output-water level" curve, the water level Lx corresponding to the maximum output Qx before the output change and the water level Li corresponding to the minimum output Qi after the output change are obtained. The difference between the two is used as the predicted water level change H.

[0141] 7) The early warning information shall publish the estimated arrival time T and the amount of water level change H, both of which shall be updated periodically. For example, the delay time T shall be updated every minute, and the time displayed each time is 1 minute less than the previous update, until the displayed value becomes 0. The amount of water level change H shall be updated every 2 minutes, and the value displayed each time is the difference between the water level Lx corresponding to the maximum output Qx before the power output change and the water level Ly corresponding to the current output value Q of the power station.

[0142] 8) The timer starts from the first issuance of the warning information. Real-time water level data L at the warning location is acquired every second and stored for a duration of t0.

[0143] 9) Compare the difference between the water level value Lt0 at the warning location before time t0 and the current water level value L. If the difference exceeds the limit h0, stop timing. The recorded duration is the actual water level change delay time Ts at the warning location under this sudden change in power output.

[0144] 10) The power plant output (Qx) before the warning was triggered, the output change value (q), and the actual water level change delay time (Ts) were used as a new set of data to optimize the warning model. The real data obtained from each warning can be used to optimize the original warning model, making the predicted water level change delay time (T) more accurate.

[0145] In summary, the technical solution of this invention uses changes in power plant output as the trigger signal for water level change early warning information. Utilizing the principle that the speed of water level change transmission is much slower than the speed of information propagation, it issues early warnings of water level changes to the warning location in advance. The advance warning time is approximately the lag time between the power output change and the resulting water level change at the warning location. This time is related to the propagation speed of water level changes; theoretically, the greater the fluctuation in power plant output, the faster the propagation speed of water level changes.

[0146] Specifically, the relationship between changes in power plant output data and the rate of change in water level (i.e., the lag time) is used to predict the expected arrival time T of downstream water level changes. This, combined with the known power plant output-downstream water level relationship curve, yields the predicted downstream water level change value H. These two sets of early warning data are crucial for downstream locations to take appropriate countermeasures, enabling them to prepare in advance. The principle that the speed of information transmission for sudden changes in power output is greater than that for changes in water level can be utilized to achieve early warning of water level changes at the warning location.

[0147] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early warning of downstream water level in a power station based on sudden changes in power output, characterized in that, include: Acquire real-time power output data of the power plant and real-time water level data of early warning points; When the power plant output change data meets the first preset condition, an early warning is triggered and an early warning message is issued. The early warning information includes the water level change delay time and / or the water level change amplitude; the determination of the water level change delay time includes: Based on the real-time water level data, and according to the early warning model, the target time point at which the real-time water level data meets the second preset condition is determined. The difference between the target time point and the warning time point when the warning information was issued is used as the water level change delay time. The generation of the early warning model includes: Obtain basic relationship data; Based on the aforementioned basic relationship data and a preset fitting method, a fitting function is generated as the initial early warning model. Based on measured data, the initial early warning model is iteratively optimized to obtain the early warning model; The underlying relational data includes: Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the first change value, and the start time of the sudden change in output is recorded until the first time when the water level change at the warning location reaches the preset change value. Under the condition that the power plant output is stable at the first preset value, within the first preset time period, the power plant output is reduced by the second change value, and the start time of the sudden change in output is recorded until the second time when the water level change at the warning location reaches the preset change value. Under the condition that the power plant output is stable at the second preset value, within the first preset time period, the power plant output is reduced by the first change value, and the time of the sudden change in output is recorded until the third time when the water level change at the warning location reaches the preset change value.

2. The method for early warning of downstream water level in a power station based on sudden changes in power output, as described in claim 1, is characterized in that, The measured data includes: The actual measured power plant output data before the warning is triggered, the actual power plant output change data, and the actual water level change delay time, wherein the actual water level change delay time is the difference between the actual time point when the real-time water level data meets the second preset condition and the warning time point when the warning information is issued.

3. The method for early warning of downstream water level in a power station based on sudden changes in power output, as described in claim 1, is characterized in that... The triggering of the warning includes: The early warning information includes the estimated arrival time T of the water level change and the amount of water level change H, both of which are updated regularly. Among them, the estimated time of water level change is displayed one minute less than the previous refresh time each time it is refreshed, until the displayed value becomes 0; The water level change H is refreshed every 2 minutes. Each time it is refreshed, the value displayed is the difference between the water level corresponding to the maximum output value before the power output change and the water level corresponding to the current power output value of the power station. The timer starts from the first release of the warning information, and real-time water level data of the warning point is acquired and stored every second for a storage period of t0; Compare the water level value Lt0 at the warning location before time t0 with the current real-time water level value. If the difference exceeds the limit h0, stop timing. The recorded duration is the actual water level change delay time under this sudden change in output.

4. The method for early warning of downstream water level in a power station based on sudden changes in power output, as described in claim 1, is characterized in that... The early warning model is a machine learning model; The step of determining the target time point where the real-time water level data meets the second preset condition based on the early warning model includes: Based on the real-time water level data and the early warning water level data, the target time point is determined through the early warning model.

5. The method for early warning of downstream water level in a power station based on sudden changes in power output, as described in claim 1, is characterized in that... The first preset condition includes: the power plant output change data exceeds 400MW within 10 seconds.

6. A power station downstream water level early warning system based on sudden output changes, characterized in that, The method for implementing the downstream water level early warning method for power plants based on sudden output changes as described in any one of claims 1 to 5 includes: The acquisition module is used to acquire real-time power output data of the power plant and real-time water level data of the early warning points; The early warning module is used to trigger an early warning and issue an early warning message when the power output change data of the power plant meets the first preset condition; The warning information includes the water level change delay time and / or the water level change amplitude; the determination of the water level change delay time includes: Based on the real-time water level data, and according to the early warning model, the target time point at which the real-time water level data meets the second preset condition is determined. The difference between the target time point and the warning time point when the warning information was issued is used as the water level change delay time.

7. A power station downstream water level early warning device based on sudden output changes, characterized in that, The device includes at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is used to execute at least some of the computer instructions to implement the power plant downstream water level early warning method based on power output mutation as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing computer instructions that, when executed by a processor, implement the downstream water level early warning method for a power station based on sudden output changes as described in any one of claims 1 to 5.

Images