CN-122020455-A - Small hydropower station hydrologic data real-time prediction and unit start-stop control method

CN122020455ACN 122020455 ACN122020455 ACN 122020455ACN-122020455-A

Abstract

The invention relates to the technical field of intelligent water conservancy, in particular to a method for predicting hydrologic data of small hydropower stations in real time and controlling start and stop of a unit. The method comprises the steps of collecting small hydropower station hydrologic data, carrying out abnormal value detection and filtering processing, outputting basic hydrologic data, obtaining historical hydrologic records, constructing a front pool water level dynamic prediction model according to the historical hydrologic records and the basic hydrologic data, generating a front pool water level prediction result, carrying out water level parameter correction on the front pool water level prediction result by utilizing the basic hydrologic data, outputting a front pool water level correction prediction result and prediction uncertainty, obtaining small hydropower station water head scheduling data, and establishing start-stop triggering conditions based on the front pool water level correction prediction result, the prediction uncertainty and the small hydropower station water head scheduling data. The invention solves the problem of unstable power generation caused by strong randomness of water supply and poor regulating capability of small water power based on the water power control technology, and remarkably improves the operation efficiency and the automation level.

Inventors

JIANG XIANQUN
XU YAN
SUN LI
WEN TAO
LEI JIAXIN
Wu Weidie
XIONG JIAYI

Assignees

广东华南水电高新技术开发有限公司

Dates

Publication Date: 20260512
Application Date: 20260113

Claims (10)

1. A method for real-time prediction of small hydropower station hydrologic data and control of start-stop of a unit is characterized by comprising the following steps: s1, acquiring small hydropower station hydrologic data, detecting abnormal values, filtering and outputting basic hydrologic data; Step S2, acquiring a historical hydrological record, constructing a dynamic prediction model of the front pool water level according to the historical hydrological record and basic hydrological data, and generating a front pool water level prediction result; Step S3, acquiring small water power head dispatching data, and establishing start-stop triggering conditions based on a front pool water level correction prediction result, prediction uncertainty and the small water power head dispatching data; And S4, determining a unit coordination priority according to the start-stop triggering condition, executing load protection based on unit coordination priority data, recording a unit operation result, transmitting the unit operation result to a front pool water level dynamic prediction model, and iterating to output a small water power start-stop control report.
2. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 1, wherein in step S2, a dynamic prediction model of a front pool water level is constructed according to historical hydrologic records and basic hydrologic data, and generating a front pool water level prediction result comprises: Constructing a front pool water level dynamic prediction model according to the historical hydrologic records and the basic hydrologic data, wherein the front pool water level dynamic prediction model comprises an input normalization layer, a hydrologic feature extraction layer and a water level time sequence prediction layer; The input normalization layer is used for carrying out normalization processing on the historical hydrologic records and the basic hydrologic data to obtain normalized hydrologic data, and the hydrologic feature extraction layer is used for carrying out activation function mapping on the normalized hydrologic data to obtain a multi-scale activation feature map; The water level time sequence prediction layer is used for carrying out linear regression operation on the pooled data to generate a pre-pool water level prediction result.
3. The method for real-time prediction of small hydropower station hydrologic data and control of start-stop of a machine set according to claim 1, wherein in step S2, the water level parameter correction is performed on the front pool water level prediction result by using the basic hydrologic data, and outputting the front pool water level correction prediction result and the prediction uncertainty comprises: The method comprises the steps of extracting real-time water level data in basic hydrological data, calculating a water level residual sequence according to the real-time water level data and a forebay water level prediction result, and determining a prediction fusion weight based on the water level residual sequence; And adjusting the dynamic prediction model of the front pool water level according to the prediction fusion weight, outputting a front pool water level correction prediction result, and calculating the prediction uncertainty according to the front pool water level correction prediction result.
4. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 1, wherein step S3 comprises: the method comprises the steps of acquiring small water power head dispatching data, extracting a water head sequence of the small water power head dispatching data, integrating a front pool water level correction prediction result and a prediction uncertainty into a prediction mapping sequence according to time dimension; generating a start-stop threshold sequence based on the predicted mapping sequence and the water head sequence, and establishing start-stop trigger conditions according to the start-stop threshold sequence.
5. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 4, wherein generating a start-stop threshold sequence based on a prediction mapping sequence and a water head sequence comprises: extracting a predictive mapping segment based on the predictive mapping sequence; extracting a water head matching section based on the water head sequence, and calculating a water head difference value according to the prediction mapping section and the water head matching section; And generating a start-stop threshold sequence according to the forward trigger fragment and the reverse trigger fragment.
6. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 5, wherein the step of screening the forward trigger segment and the reverse trigger segment according to the water head difference value comprises the steps of: Setting a forward water head threshold and a reverse water head threshold, wherein the forward water head threshold is larger than zero, and the reverse water head threshold is smaller than zero; starting from the forward trigger starting point, continuously tracking the subsequent water head difference value until the water head difference value falls back below the forward water head threshold value, and marking the segment as a forward trigger segment; And starting from the reverse triggering starting point, continuously tracking the subsequent water head difference value until the water head difference value rises above the reverse water head threshold value, and marking the segment as a reverse triggering segment.
7. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 1, wherein step S4 comprises: screening available units according to the start-stop triggering condition, and calculating the load distribution quantity of the available units; And transmitting the unit operation result to a front pool water level dynamic prediction model, and iterating to output a small water power on-off control report.
8. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 7, wherein the steps of screening available units according to start-stop triggering conditions and calculating load distribution amounts of the available units include: the method comprises the steps of screening available units according to start-stop triggering conditions, recording unit load degrees of the available units, determining unit load trend according to the unit load degrees, calculating load impact quantity based on the unit load trend, and determining load distribution quantity of the available units according to the load impact quantity.
9. The method for real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 7, wherein performing load protection based on unit coordination priority data and recording unit operation results comprises: Performing load protection simulation based on the unit coordination priority data, and synchronously recording load protection data, calculating the opening of a unit guide vane according to the load protection data, calculating the water flow speed according to the opening of the unit guide vane, and evaluating the radial hydrodynamic force of the rotating wheel according to the water flow speed; Generating a bearing load sequence according to radial hydrodynamic force of the rotating wheel, counting load fluctuation values according to the generated bearing load sequence, and taking the load fluctuation values as a unit operation result.
10. The method of real-time prediction and unit start-stop control of small hydropower station hydrologic data according to claim 9, wherein executing load protection simulation based on unit coordination priority data and synchronously recording load protection data comprises: Extracting a priority segmentation sequence according to the unit coordination priority data, and carrying out load protection simulation according to the priority segmentation sequence so as to determine a real-time load state; The method comprises the steps of comparing a real-time load state with a preset load protection target section by section to obtain a load protection comparison result, screening triggerable protection fragments based on the load protection comparison result, and recording load protection data according to the triggerable protection fragments.

Description

Small hydropower station hydrologic data real-time prediction and unit start-stop control method Technical Field The invention relates to the technical field of intelligent water conservancy, in particular to a method for predicting hydrologic data of small hydropower stations in real time and controlling start and stop of a unit. Background In the existing small hydropower station operation management and unit start-stop control technology, manual experience judgment or a rule control mode based on a fixed threshold value is generally relied on to respond to hydrologic changes, for example, unit start-stop conditions are manually set according to real-time water inflow, front pool water level or experience water head intervals. The mode can maintain basic operation under the scenes of relatively stable hydrologic conditions, slow water supply change or less unit quantity, but under the conditions of frequent water supply fluctuation and obvious randomness and hysteresis of hydrologic change of small hydropower stations in mountain areas, the prior art is difficult to timely and accurately guide the start-stop decision of the unit due to lack of the capability of predicting the hydrologic evolution trend in advance. The method is mainly characterized in that a first method is focused on current hydrologic state judgment, combined modeling of historical hydrologic data and real-time data is lacked, dynamic prediction of front pool water level change is difficult, start-stop decision delay is easily caused, a second method is mainly used for start-stop control, fixed water level or water head threshold value is not fully considered, prediction uncertainty and water head scheduling constraint are not fully considered, frequent start-stop or load impact of a unit is easily caused, and third, effective linkage is lacked between a unit start-stop process and load protection and operation result evaluation, unit operation state cannot be reversely fed back to a hydrologic prediction model for correction, a closed-loop optimization mechanism is lacked in an overall control flow, and requirements of safety, stability and fine operation management of small hydropower are difficult to meet. Disclosure of Invention Based on the above, the present invention is necessary to provide a method for real-time prediction of small hydropower station hydrologic data and control of start-stop of a machine set, so as to solve at least one of the above technical problems. In order to achieve the purpose, the method for predicting the hydrologic data of the small hydropower station in real time and controlling the start and stop of the unit comprises the following steps: s1, acquiring small hydropower station hydrologic data, detecting abnormal values, filtering and outputting basic hydrologic data; Step S2, acquiring a historical hydrological record, constructing a dynamic prediction model of the front pool water level according to the historical hydrological record and basic hydrological data, and generating a front pool water level prediction result; Step S3, acquiring small water power head dispatching data, and establishing start-stop triggering conditions based on a front pool water level correction prediction result, prediction uncertainty and the small water power head dispatching data; And S4, determining a unit coordination priority according to the start-stop triggering condition, executing load protection based on unit coordination priority data, recording a unit operation result, transmitting the unit operation result to a front pool water level dynamic prediction model, and iterating to output a small water power start-stop control report. The beneficial effects of the invention are as follows: (1) By means of joint modeling of historical hydrologic records and real-time hydrologic data, a dynamic prediction model of the water level of the front pool is built, water level parameter correction and prediction uncertainty assessment are introduced, advanced sensing and fine prediction of the water level change trend of the front pool are achieved, time sequence continuity and reliability of hydrologic prediction results are improved, and a stable data base is provided for unit start-stop decision. (2) In the unit start-stop control process, a start-stop threshold sequence is generated based on a front pool water level correction prediction result, prediction uncertainty and water head scheduling data, and the frequent start-stop problem caused by a single fixed threshold is avoided through a segmented screening mechanism of a forward trigger segment and a reverse trigger segment, so that the smoothness and adaptability of a unit start-stop decision are enhanced. (3) The method combines the coordination priority, load protection simulation and radial hydrodynamic force evaluation of the rotating wheel, quantitatively analyzes load change and load fluctuation in the running process of the unit, and feeds back running result