New energy station intelligent micrometeorological monitoring device suitable for extreme weather and power prediction method

Abstract

The invention discloses a new energy station intelligent micrometeorological monitoring device suitable for extreme weather and a power prediction method, and relates to the technical field of new energy station meteorological monitoring and power prediction. The device integrates a plurality of meteorological sensors, combines satellite meteorological data and ground meteorological station data, improves meteorological monitoring accuracy through data fusion, and gives 8-15-day long-time scale weather forecast. According to the method, based on meteorological data collected in real time and historical meteorological data of a new energy station, a long-short term memory network model and a segmentation multi-target risk method are adopted to carry out risk assessment on extreme weather, and an extreme weather early warning signal is given. A deep neural network is utilized to establish a power nonlinear prediction model of a new energy station, and a higher-precision power prediction result is given through a random forest dynamic correction algorithm. The meteorological and generated power prediction accuracy of the new energy station is improved, and the operation safety of the new energy station is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of micro-meteorological monitoring and power prediction of new energy stations, and in particular to a long-term micro-meteorological monitoring and power prediction method for new energy stations under extreme weather conditions. Background Art

[0002] As the global energy structure transforms towards a low-carbon and green direction, the construction and application of new energy stations have been promoted. However, the stable operation of new energy stations faces many challenges. The accuracy of micro-meteorological monitoring and power generation prediction is an important part of ensuring the efficient operation of new energy stations.

[0003] The existing power generation prediction methods for new energy stations are insufficient in mining the spatiotemporal correlation information in regional numerical weather forecasts, resulting in large prediction errors. The micro-meteorological monitoring devices of new energy stations perform poorly in terms of anti-interference ability in the external environment, especially under extreme weather conditions such as strong winds, heavy rains, and high temperatures, the accuracy of meteorological forecasts is insufficient. In addition, the data processing capacity of existing equipment is limited, and it is difficult to fully utilize the multi-source information of meteorological sensors, satellite meteorological data, and ground meteorological station data. The meteorological forecast time scale is short and can only meet the short-term forecast needs of 1 to 3 days.

[0004] The efficient operation of new energy stations not only depends on accurate meteorological monitoring, but also requires reliable power generation prediction. Scientific micro-meteorological monitoring and power prediction can effectively optimize station operation, reduce the probability of wind and solar power abandonment, station equipment damage, etc., improve the utilization rate of new energy, and ensure the safety of equipment and personnel by early warning of extreme weather. Therefore, designing an intelligent micro-meteorological monitoring device and power prediction method suitable for extreme weather has become a key technical problem that needs to be solved in the field of new energy technology. Summary of the invention

[0005] The purpose of this application is to provide an intelligent micro-meteorological monitoring device and power prediction method for new energy stations suitable for extreme weather conditions, so as to solve the problems of insufficient micro-meteorological monitoring accuracy, short meteorological prediction time scale, inaccurate extreme weather risk assessment and large power prediction error in the prior art. This application realizes the intelligent operation and efficient management of new energy stations through multi-source data fusion technology, collaborative prediction method of numerical model and deep learning model, and dynamic extreme weather risk assessment mechanism.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides an intelligent micro-meteorological monitoring device for a new energy power station. Based on the complex meteorological conditions of the new energy power station, the present application proposes an intelligent micro-meteorological monitoring device, which includes a meteorological data acquisition module, a multi-source data fusion module, and a data transmission module. Through the optimization of hardware design and the integration of advanced algorithms, the efficient combination of real-time meteorological monitoring and data processing is achieved.

[0008] Meteorological data acquisition module: Integrate wind speed and direction sensors to monitor the wind speed V and direction in real time. The data is used for the calculation of wind power. Integrate temperature and humidity sensors to collect the ambient temperature and relative humidity, which are used to judge the atmospheric stability and radiation energy. Integrate a pressure sensor to measure the pressure P and provide information on the change in atmospheric pressure. Integrate a light sensor to capture the solar radiation intensity G t , which is used for photovoltaic power prediction. Construct a three-dimensional sensor network to cover the multi-level spatial area of the new energy power station. Adopt a redundant layout and a real-time diagnosis mechanism to enhance the fault tolerance and reliability of the acquisition system. Even in the case of sensor node failures or external interference, the continuity and stability of data acquisition can still be ensured.

[0009] Multi-source data fusion module: Based on the Extended Kalman Filter (EKF) algorithm, dynamically fuse sensor data, satellite meteorological data, and ground meteorological station data to optimize the meteorological state estimation. Through the iterative process of state prediction and observation correction, generate high-precision meteorological data, providing a solid data foundation for subsequent meteorological prediction and risk assessment. State monitoring and state update:

[0010] x k = F k x k-1 + B k u k + w k ;

[0011] x' k = x k + K k (z k - H k x k )

[0012] where x k is the previous state vector, representing meteorological parameters (such as wind speed V, temperature T, humidity H, etc.), F k is the state transition matrix, representing the evolution law of the meteorological state over time, u k is the control input, representing external influences, B k is the control matrix, mapping the input quantity to the state, w k is the process noise, representing unmodeled random disturbances, z k is the observation value vector, representing the meteorological data collected by the sensors, Hk is the observation matrix, representing the linear relationship between observations and states, K k is the Kalman gain, which is used to balance the weights of the predicted value and the observed value.

[0013] Data transmission module: Utilize the Narrowband Internet of Things (NB-IoT) technology to achieve efficient and low-power data transmission, meeting the long-term monitoring requirements of new energy power stations.

[0014] In a second aspect, the present application provides a long-term meteorological prediction method. To overcome the problems of short meteorological prediction time range and insufficient accuracy in the prior art, the present application proposes a long-term meteorological prediction method combining a numerical weather prediction model (WRF) and a deep learning model (LSTM). The WRF model is based on the atmospheric dynamics and thermodynamics equations, numerically simulating meteorological variables such as wind speed, temperature, humidity, and light intensity to generate basic prediction data. The LSTM model is used to correct the systematic error. For the error in the output result of the WRF model, the LSTM dynamically adjusts the prediction result by learning the time series characteristics of meteorological variables.

[0015] In a third aspect, the present application provides an extreme weather risk assessment method. Aiming at the uncertainty of extreme weather events, the present application proposes a risk assessment method combining Bayesian inference and random forest.

[0016] Bayesian inference calculates the posterior probability of extreme weather through the following formula:

[0017]

[0018] P(E|D) is the posterior probability, representing the likelihood of the occurrence of extreme weather event E under the observed data D. P(E) is the prior probability, which is statistically obtained based on historical extreme weather data. P(D|E) is the conditional probability, representing the distribution of the observed data D when E occurs.

[0019] Use the random forest to classify meteorological features. Based on the PMRM segmentation multi-objective risk method, the extreme weather risk is divided into three levels: low risk R L , medium risk R M and high risk R H The decision tree model can dynamically adjust the weights according to real-time meteorological variables, improving the accuracy of the evaluation result. Through the linkage with the meteorological prediction module, the risk assessment result is updated in real time to ensure the timeliness and adaptability of the evaluation. In the high-risk level, the wind turbine executes load reduction or shutdown protection, and the photovoltaic power station adjusts the component inclination. In the medium-risk level, the station equipment remains in a vigilant state while continuing to optimize the power output. In the low-risk level, the equipment operates normally and continuously monitors the risk changes.

[0020] Fourthly, the present application provides a power prediction method for new energy power stations. Based on meteorological monitoring data and prediction data, the present application proposes a power prediction method combining a kinetic model and a deep learning model, which is applicable to wind farms and photovoltaic power stations. When calculating the basic power by the kinetic model, the power output P of the wind farm w is calculated by the following formula:

[0021]

[0022] where ρ is the air density, related to temperature and air pressure, P is the air pressure, R is the air gas constant, and A is the swept area of the wind turbine. C p is the power coefficient, which is the efficiency parameter of the wind turbine. The wind speed V comes from the wind speed sensor of the micro-meteorological monitoring device.

[0023] The power output of the photovoltaic power station is calculated by the following formula:

[0024] P pv = G t ηS

[0025] where G t is the solar radiation intensity per unit time, provided by the light sensor. η is the conversion efficiency of the photovoltaic module, depending on the module material, temperature and light conditions. S is the effective area of the photovoltaic module.

[0026] The kinetic model depends on meteorological parameters, but does not consider the correlation and non-linear effects of time series, which may lead to prediction errors in complex scenarios. To overcome the deficiencies of the kinetic model in complex scenarios, the present application introduces a long short-term memory network (LSTM) model to correct the power prediction error by using the time series characteristics of micro-meteorological data. The input includes meteorological data at the current moment and historical moments, including the initial predicted power generation P 0 , wind speed V t , light intensity G t , temperature T t , and humidity H t . The output is the corrected power prediction value P'.

[0027] Fifthly, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the micro-meteorological monitoring and power prediction methods for new energy power stations described in any one of the above.

[0028] Sixthly, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the micro-meteorological monitoring and power prediction methods for new energy power stations described in any one of the above.

[0029] According to the specific embodiments provided by the present application, the following technical effects are disclosed in the present application:

[0030] By designing an intelligent micro-meteorological monitoring device, a long-term meteorological prediction method, an extreme weather risk assessment method, and a power prediction method, the present application comprehensively improves the operation efficiency and safety of new energy power stations under complex meteorological conditions. First, the intelligent micro-meteorological monitoring device realizes high-precision and real-time meteorological data collection and dynamic update within the power station area through a three-dimensional sensor network and a multi-source data fusion module. Based on the extended Kalman filtering technology, the data from sensors, satellites, and ground stations are fused, effectively reducing the monitoring error and providing high-quality data input for subsequent predictions. Secondly, the long-term meteorological prediction method combines the advantages of the WRF model and the LSTM model, making use of the physical laws of numerical models and optimizing the non-linear modeling ability through deep learning, significantly improving the prediction accuracy of meteorological variables in the next 8 to 15 days and meeting the medium- and long-term operation planning requirements of new energy power stations. In addition, the extreme weather risk assessment method integrates Bayesian inference and the PMRM segmentation multi-objective risk model to dynamically evaluate the probability of extreme weather occurrence and decompose its multi-objective risks. The PMRM model refines the risk assessment into low risk, medium risk, and high risk by constructing sub-objective functions, and combines the random forest algorithm to achieve accurate classification of risk levels, ensuring the scientificity and timeliness of equipment protection strategies. In terms of power prediction, the present application combines a kinetic model and an LSTM model, using real-time meteorological data and historical power data, not only significantly improving the accuracy of wind power and photovoltaic power predictions, but also dynamically adjusting the power output strategy under extreme weather conditions to avoid equipment damage and power generation interruption caused by sudden weather. Through the above technical solutions, the present application comprehensively improves the operation safety, economy, and stability of new energy power stations, and has broad application prospects and significant technical advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0032] Figure 1 Schematic diagram of the micro-meteorological forecasting and power prediction method for a new energy power station provided by an embodiment of the present application; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0034] The purpose of the present application is to provide an intelligent micro-meteorological monitoring device and power prediction method for new energy power stations applicable to extreme weather, aiming to comprehensively consider multi-source meteorological data, improve the time scale and accuracy of meteorological forecasts, evaluate the risks of new energy power stations under extreme weather, and improve the accuracy of power prediction for new energy power stations.

[0035] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0036] In an exemplary embodiment, as Figure 1 , the micro-meteorological monitoring device and power prediction method in this embodiment include:

[0037] This embodiment provides an intelligent micro-meteorological monitoring and power prediction method for new energy power stations, which combines the fusion processing of multi-source meteorological data, long-term meteorological prediction, extreme weather risk assessment, and power prediction and operation strategy adjustment, realizing the efficient management and safety protection of power station operation.

[0038] Step 1: Meteorological data collection

[0039] Real-time collect the meteorological data set D from the multi-source meteorological monitoring devices arranged in the new energy power station w , the anemometer and wind vane sensors, temperature and humidity sensors, barometric pressure sensors, and light sensors connected to the meteorological monitoring devices can accurately capture key meteorological parameters such as wind speed, wind direction, temperature and humidity, barometric pressure, and light intensity. The collection range covers multiple spatial levels of the power station, and through reasonable sensor layout and redundancy design, the continuity and reliability of data collection are ensured. The collected data is transmitted to the data processing module through NB-IoT technology, providing real-time input for subsequent data processing and prediction.

[0040] Step 2: Multi-source data fusion

[0041] Fuse the collected sensor data with satellite meteorological data and ground meteorological station data to generate high-precision meteorological monitoring data D p, the fusion process adopts the extended Kalman filtering method, which can effectively correct the errors caused by sensor noise or data loss, and improve the accuracy of meteorological state estimation. By dynamically adjusting the model parameters, the fused data combines the advantages of multi-source information, not only has the ability to distinguish local details, but also covers a large-scale meteorological background, providing high-quality data support for the operation of the station.

[0042] Step 3: Long-term meteorological prediction

[0043] Based on the fused meteorological data D p , use the numerical weather prediction model to conduct basic meteorological prediction and generate meteorological trend information W for a future period of time. p . The numerical weather prediction model is based on the basic laws of atmospheric dynamics and thermodynamics and can provide meteorological prediction results with relatively high scientificity. However, in order to further improve the prediction time scale and accuracy, this method combines historical meteorological data and uses the long short-term memory network (LSTM) model to correct the output results of the numerical weather prediction model to generate high-precision meteorological prediction data W. a . The LSTM model can effectively capture the non-linear relationship and time series characteristics of meteorological variables, significantly reduce the cumulative error in long-term prediction, and achieve accurate meteorological prediction for the next 8 to 15 days.

[0044] Step 4: Extreme weather risk assessment

[0045] Based on the high-precision meteorological prediction data W a , construct an extreme weather risk assessment module to dynamically evaluate the probability of extreme weather occurrence and its risk level in the station area. The assessment process combines Bayesian inference to calculate the probability of extreme weather occurrence, and uses the PMRM multi-objective risk model to refine and decompose the extreme weather risk, specifically including risk targets such as high wind speed, sudden drop in air pressure, and heavy precipitation. Subsequently, a random forest model is used to classify the risk levels to generate assessment results of low risk, medium risk, and high risk levels. The final risk assessment results can be updated in real time and provide clear decision-making basis for the operation strategy of the station.

[0046] Step 5: Power prediction

[0047] Use the meteorological prediction data W a and historical power data to construct a power prediction model and calculate the power generation output of the station. First, calculate the basic power output value P of the wind farm and the photovoltaic power station based on the meteorological dynamics model. b . The dynamics model can provide physically reasonable power prediction results based on real-time meteorological data. However, in order to adapt to the non-linear changes in complex scenarios, this method further combines the long short-term memory network model. By learning the time series characteristics of the power output, the basic power prediction value P bMake corrections to generate a more accurate power prediction result P a This method fully combines the scientific nature of the physical model and the flexibility of the deep learning model, effectively improving the accuracy of power prediction.

[0048] Step 6: Adjustment and early warning of the operation strategy of the new energy power station

[0049] Based on the extreme weather risk assessment results and power prediction results, judge the operation risk level of the new energy power station and determine whether the early warning risk threshold is reached. When the risk level reaches the early warning threshold, the system will automatically trigger an early warning signal and adjust the operation strategy of the power station. For wind farms, choose to operate at reduced load or shutdown protection; for photovoltaic power stations, adjust the inclination angle of the components to reduce the risk of radiation overload. At the same time, combined with the real-time power prediction results, optimize the dispatching strategy of the power station to maximize power output on the premise of ensuring safety. When the risk level is lower than the early warning threshold, the power station equipment operates normally and continuously monitors the risk changes dynamically.

[0050] Through the above steps, this application realizes a complete closed loop from meteorological data collection to power prediction and risk assessment, providing strong technical support for the safe operation and efficient management of new energy power stations under complex meteorological conditions.

Claims

1. An intelligent micro-meteorological monitoring device for new energy stations suitable for extreme weather, characterized in that: include: Meteorological data acquisition module, multi-source data fusion module, meteorological forecast module, extreme weather risk assessment module and data transmission module; The meteorological data acquisition module includes a plurality of meteorological sensors, including a wind speed and direction sensor for collecting wind speed and wind direction, a temperature and humidity sensor for collecting ambient temperature and humidity, a pressure sensor for collecting air pressure, and a light sensor for collecting light intensity; The meteorological data acquisition module adopts a three-dimensional spatial arrangement to construct a three-dimensional sensor network to cover the multi-level spatial areas of the new energy station; The meteorological data acquisition module is connected to the multi-source data fusion module via a narrowband Internet of Things communication mode (NB-IoT); The multi-source data fusion module uses the extended Kalman filter (EKF) method to perform state estimation and error correction on multi-source meteorological data, fusing multi-source information from meteorological sensors, satellite meteorological data and ground meteorological station data; The weather forecast module generates basic weather forecast results for the next 8 to 15 days based on the numerical weather forecast model (WRF), and corrects the basic forecast results in combination with the deep learning model long short-term memory network (LSTM); The extreme weather risk assessment module calculates the probability of extreme weather occurrence through Bayesian inference (BI), calculates the extreme weather risk value based on the segmentation multi-objective risk method (PMRM), and uses random forest (RF) to classify meteorological characteristics to generate extreme weather risk levels and warning signals; The data transmission module sends the revised weather forecast data, extreme weather risk assessment results and power prediction results to the new energy station control center through NB-IoT to optimize the station's operation strategy.

2. A method for predicting power of new energy stations suitable for extreme weather, characterized in that: The power prediction method performs power prediction based on a hybrid model, specifically comprising: The basic power output is calculated using the meteorological dynamics model. The deep learning model LSTM is used to correct the error of the basic power prediction results by combining the meteorological forecast data with the historical power generation data. The revised power forecast results are optimized again through a dynamic optimization algorithm to generate high-precision power forecast data; the final power forecast data is used for station operation decision-making and power adjustment; When the predicted power exceeds the safe power generation limit of the new energy unit or the weather at the site is extreme, a command to limit the site's power generation power and unit operation will be issued.

3. The intelligent micro-meteorological monitoring device for new energy stations according to claim 1 is characterized in that: The meteorological data acquisition module adopts the following measures to improve the reliability and accuracy of data acquisition: Distribute multiple meteorological sensors in the horizontal and vertical directions to build a three-dimensional sensor network covering the entire station area; configure backup sensor nodes through redundant arrangement to improve system fault tolerance; monitor the operating status of sensors in real time, and supplement missing data through redundant nodes when a failure occurs.

4. The intelligent micro-meteorological monitoring device for new energy stations according to claim 1 or 2, characterized in that: The multi-source data fusion module uses the following steps to improve the accuracy of meteorological state estimation: Based on the extended Kalman filter (EKF) method, the state of the collected real-time meteorological data and multi-source data is estimated; the macro information from satellite meteorological data and the local details of ground meteorological station data are integrated; the Kalman gain is dynamically adjusted to optimize the state estimation result according to the noise characteristics of the sensor signal; and high-precision meteorological state data is generated through the filtered multi-source data for use by the meteorological forecast module.

5. The intelligent micro-meteorological monitoring device for new energy stations according to claims 1 to 3 is characterized in that: The weather forecast module comprises: The numerical weather prediction model WRF, based on multi-source meteorological data, is used to predict changes in meteorological variables within the next 8 to 15 days, including wind speed, wind direction, temperature and humidity, air pressure, and light intensity; The LSTM deep learning model based on historical meteorological data can correct the error of the output results of the numerical weather forecast model by learning the time series trend and nonlinear relationship of meteorological variables, and generate high-precision meteorological forecast data; Update the input data of the extreme weather risk assessment module according to the meteorological forecast results.

6. The intelligent micro-meteorological monitoring device for new energy stations according to claims 1 to 4, characterized in that: The extreme weather risk assessment module adopts the following steps: Use Bayesian inference BI to dynamically calculate the probability of extreme weather; Based on the probability results and the extreme weather risk values ​​calculated by the segmentation multi-objective risk method, the random forest RF is used to classify extreme weather risks into three levels: low risk, medium risk, and high risk; Generate extreme weather warning signals based on risk levels and send the signals to the data transmission module.

7. The power prediction method according to claim 2, characterized in that: The meteorological dynamical model includes: Wind speed-power curve models for wind farms, which generate wind power output by calculating variables such as wind speed, air density, swept area, and power factor; The photovoltaic radiation conversion model used in photovoltaic power plants generates photovoltaic power output by calculating light intensity, photovoltaic module efficiency and radiation area.

8. The power prediction method according to claims 2 and 7, characterized in that: The deep learning model LSTM optimizes the power prediction results through the following steps: Capture the time series characteristics of power output and analyze the historical trends of meteorological variables; Modeling the nonlinear relationship between meteorological variables and power output to optimize the prediction error of the kinetic model; The power output in complex scenarios is corrected through a multi-layer neural network structure to improve prediction accuracy.

9. The intelligent micro-meteorological monitoring device for new energy stations according to claims 1 to 8, characterized in that: The data transmission module sends the revised weather forecast data, extreme weather risk assessment results and power prediction results to the site control center through the narrowband Internet of Things communication mode NB-IoT; at the same time, it supports remote updating of equipment parameters and algorithm models.

Images