"a system for accurately determining long-term forecast and method thereof"

Abstract

The present invention provides a system and a method to accurately determine the forecasts of Irradiation (GHI) & Wind Speed (at hub height of wind turbine) The present invention comprises an anemometer (100), a controlling means 1 (102), a weather monitoring station (104), a controlling means 2 (106), a communicating means (108), a storage means (110) and a processing means (112). Further, the present invention accurately determines the forecast for the next 1 to 25 years with up to 96% accurate forecast of wind speed and up to and up to 98% accurate forecast of global solar horizontal radiation.

Description

[0001] “A SYSTEM FOR ACCURATELY DETERMINING LONG-TERM FORECAST AND METHOD THEREOF”

[0002] FIELD OF THE INVENTION:

[0003] The present invention relates to field of weather forecasting. In particular, the present invention provides a system and a method to accurately determine the forecasts of Irradiation (GHI) & Wind Speed (at hub height of wind turbine) for the next 1 year to 25 years. Further, the present invention is used to form critical & essential analysis for solar & wind farm design and commercials.

[0004] BACKGROUND OF THE INVENTION:

[0005] Accurate Prediction of Wind and Solar Resources help in deriving optimum bid and energy pricing for given capacity of Renewable plant. Climate (long-term average of weather conditions in a particular region or across the globe) is characterized by meteorological variables such as temperature, precipitation, wind speed, and humidity. The Earth's climate has been changing naturally over millions of years due to various factors, including variations in the Earth's orbit and solar radiation.

[0006] However, in recent times, according to the recent IPCC report, the climate is changing due to global warming (increase in Earth's average surface temperature) caused by human induced emission particularly the burning of fossil fuels, have been the primary cause of climate change. Renewable wind & solar energy industries across Indian subcontinent saw an unprecedented drop in energy production in the most recent years. These observations are critical as multiple academic studies also suggest a declining trend in the wind & solar power potential availability in India due to these climatic activities.

[0007] This in turn directly affects energy production as well as the physical resilience of current and future energy infrastructure. Earlier resource estimation process doesn’t take into consideration the climate change impact and detrimental effect of Black Carbon concentrations on Solar and wind resource estimations along with other socioeconomic factors and local climatic aspects. These uncertainties in resource estimation are negatively impacting project financing and its commercial viability.

[0008] Earlier, instead of forecasting, resource was estimated based on the average historical values of solar irradiation and wind speeds for the previous 15-20 years and production yield estimates for plant life (next 25 years) are derived from these average numbers. Earlier methodology assumes that average historical long term resource values will remain constant for the lifetime of plant which is 20-25 years in the future and due to these assumptions YoY deviations in solar and wind resources were observed with a yearly aggregated resource estimation accuracy of around 94% for solar and 90% for wind. The major portion of inaccuracies in the earlier method comprises of lack of inclusion of climate change patterns, Socio Economic changes over the region, long term effects of change in concentration of pollutants in atmosphere. Present system is a novel method which projects 1 year to 25 years in future and take care of all the shortcomings of the earlier method resulting in up-to 4% increase in annual solar resource forecast accuracy and up-to 6% in annual wind forecast accuracy over earlier method. Also, unlike earlier method it does not require at least 2-3 years of historical actual site measured data for estimation of wind and solar resources.

[0009] Renewable business is by nature a variable resource business, hence, accurate prediction of solar and wind resources for the next one year and 25 years is crucial for solar / wind / hybrid (with and without storage) plants. It helps in deriving optimum bid and energy pricing for a given capacity of Renewable plant as 1% change in irradiation / wind speed may lead to 0.4%-0.6% change in equity IRR for solar / wind portfolio.

[0010] Thus, there is an unmet need to develop a system and a method for accurately determining long-term forecast.

[0011] Prior art & its Disadvantages:

[0012] A U.S. patent application US10909446B2 relates to a methods and systems for generating a multi-model ensemble of global climate simulation data from a plurality of global climate simulation models (GCMs), to be used in training a neural network (NN) -based climate forecasting model, are disclosed. The methods and systems perform steps of computing a GCM validation measure for each GCM; selecting a validated subset of the GCMs, by comparing each computed GCM validation measure to a validation threshold determined based on observational historical climate data; computing a forecast skill score for each validated GCM, based on a first forecast function; selecting a validated and skillful subset of GCMs; generating one or more candidate ensembles by combining simulation data from at least two validated and skillful GCMs; computing an ensemble forecast skill score for each candidate ensemble, based on a second forecast function; and selecting a best-scored candidate ensemble. Embodiments of the present invention enable accurate climate forecasting without the need to run new dynamical global climate simulations on supercomputers.

[0013] However, said prior art invention provides forecasting for shorter duration and specifically for wind forecasting (6 months) without the need to run new dynamical global climate simulations on supercomputers and thus, fails to provide a system and method to accurately determine long-term forecast.

[0014] Another patent application JP2023016742A relates to a computer- implemented method of forecasting weather events includes generating a training model via artificial intelligence by a computer processor. The training model is based on climate data processed by a variation auto encoder. Geographic locations are selected for climate studies. Historical weather measurements associated with the selected geographic location are retrieved from the knowledge climate database. Obtained historical weather measurements are processed using a training model. The training model may receive threshold parameters that define weather extremes. Extreme values are based on weather intensity data points that are farther from the norm than near the norm. Synthetic weather data is generated for the selected location, and the synthetic weather data predicts weather events that meet extreme threshold parameters.

[0015] However, Forecasting / simulating extreme weather conditions for shorter duration and thus, fails to provide a system and method to accurately determine long-term forecast.

[0016] Yet another patent application US10996374B1 discloses a system includes at least one server implementing a weather forecast engine and in communication with a network, the server to access satellite imagery, published weather predictions, and local measured data via the network; the weather forecast engine to train regional modules using the satellite imagery, published weather predictions, and local measured data; and the weather forecast engine to apply the satellite imagery, published weather predictions, and local measured data to the trained regional modules to obtain regional forecasts. A method for forecasting a weather indicator includes receiving satellite imagery; processing the satellite imagery to generate a weather feature set; applying the weather feature set to a regional module of a weather forecast engine; and forecasting the weather indicator with the weather forecast engine. The cited document discloses a system and method are disclosed for forecasting weather features, including but not limited to wind speed and direction, air temperature, solar radiation (irradiance), atmospheric pressure, relative humidity, and more. Using artificial intelligence models (including but not limited to machine learning, deep learning, and computer vision algorithms) together with hybrid datasets (including but not limited to remote sensing / satellite images, weather forecast model outputs, and in-situ measurements, at both land and offshore sites), a flexible model is derived with respect to an objective function that also exhibits demonstrably more accurate predictions than conventional models.

[0017] However, said prior art used to forecast wind speed and direction, air temperature, solar radiation (irradiance), atmospheric pressure, relative humidity for very short duration (mostly 7 days ahead) and thus, fails to provide a system and method to accurately determine long-term forecast.

[0018] Disadvantages of the prior art:

[0019] The conventionally used system for the determination of forecast suffers from the following disadvantages:

[0020] Many of the prior arts fail to provide a system for accurately determining long-term forecast.

[0021] Many of the prior arts fail to provide a method for accurately determining long-term forecast.

[0022] Most of the prior arts fail to provide a system that accurately determines the forecast for the next 1 year to 25 years.

[0023] Most of the prior arts fail to provide a system that gives up to 96% accurate forecast of wind speed.

[0024] Most of the prior arts fail to provide a system that gives up to 98% accurate forecast of global solar horizontal radiation. Most of the prior arts fail to provide a method for accurately determining long-term forecast that results in achieving accurate estimation of Annual Energy production for the next 25 years horizon.

[0025] OBJECTS OF INVENTION:

[0026] The main object of the present invention is to provide a system for accurately determining long-term forecast.

[0027] Another object of the present invention is to provide a method for accurately determining long-term forecast.

[0028] Yet another object of the present invention is to provide a system that accurately determines the forecast for the next 1 year to 25 years.

[0029] Yet another object of the present invention is to provide a system that gives up to 96% accurate forecast of wind speed.

[0030] Yet another object of the present invention is to provide a system that gives up to 98% accurate forecast of global solar horizontal radiation.

[0031] Yet another object of the present invention is to provide a method for accurately determining long-term forecast that results in achieving accurate estimation of Annual Energy production for the next 25 years horizon.

[0032] Yet another object of the present invention is to overcome the deficiency and disadvantages of the prior arts and provide an efficient system and method for accurately determining long-term forecast. BRIEF DESCRIPTION OF DRAWINGS:

[0033] Fig 1 : Shows a block diagram of the interconnections and working of the component parts of present invention.

[0034] Fig 2 : Shows a block diagram explaining Step 1 of the present invention.

[0035] Fig 3 : Shows a block diagram explaining Step 2 of the present invention

[0036] Fig 4 : Shows a block diagram explaining Step 3 of the present invention

[0037] Fig 5 : Shows a block diagram explaining Step 4 of the present invention

[0038] Fig 6 : Shows a block diagram explaining Step 5 of the present invention

[0039] Fig 7 : Shows a block diagram explaining Step 6 of the present invention

[0040] Fig 8 : Shows a block diagram for the working of the present invention.

[0041] Reference Numerals of the component parts of present System for Accurately Determining Long-Term Forecast and Method Thereof:

[0042] SUMMARY OF INVENTION:

[0043] The present invention provides a System for accurately determining Long-Term Forecast comprising of an anemometer (100), a controlling means 1 (102), a weather monitoring station (104), a controlling means 2 (106), a communicating means (108), a storage means (110) and a processing means (112). Further, the present invention provides a method for accurately determining long-term forecast for the next 1 to 25 years. The present invention provides a system and a method that gives up to 96% accurate forecast of wind speed and up to 98% accurate forecast of global solar horizontal radiation. DETAILED DESCRIPTION OF INVENTION:

[0044] The present invention provides a system to accurately determine the forecasts of Irradiation (GHI i.e., Global Horizontal Irradiance) & Wind Speed (at hub height of wind turbine) for the next 1 year to 25 years. Further, the present invention is used to form critical & essential analysis for solar & wind farm design and commercials.

[0045] Referring to Figure 1 , the present invention of a system and a method to accurately determine forecasts comprises:

[0046] 1). Wind Turbine Speed Measurement system

[0047] The wind turbine Anemometer (100) is installed on Wind turbine at hub height. Herein the hub height is referred to as the distance from the ground to the middle of the turbine's rotor. Said anemometer (100) counts the number of rotations, or turns of the turbine, which is used to calculate wind speed and further records the real time wind speed. Said anemometer (100) is connected to a Controlling means 1 (102) using a hardware cable. Said Controlling means relays the data received from said anemometer (100) to a Controlling means 2 (106) over a Fiber-Optic communication. Said controlling means 1 (102) is preferably but not limiting to a Programmable Logic Controller (PLC) system. Said controlling means 2 (106) is preferably but not limiting to a Supervisory Control and Data Acquisition (SCAD A) . 2). Solar Plant - Weather Monitoring Station

[0048] Weather Monitoring stations i.e., WMS (104) are installed in each solar plant. Said WMS (104) provides global solar horizontal irradiation (GHI), inclined irradiation, wind speed, rain fall, ambient temperature etc.

[0049] The WMS system (104) is connected with Controlling means 2 (106) using a cable connection.

[0050] Said Controlling means 2 (106) relays these data points to Communicating means (108) and then further to a Storage means (110). Said communicating means (108) is preferably but not limiting to an Open Platform Communications (OPC) Server. Said storage means (110) is preferably but not limited to a Cloud server. This data forms one of the input databases for the processing means (112) of present system.

[0051] The processing means (112) collects the information from the storage means (110) and further analyzes the same to provide an accurate long-term weather forecast.

[0052] WORKING OF INVENTION:

[0053] The method of the present invention is categorized and explained in the following steps:

[0054] 1). Satellite Data extraction

[0055] This is in reference to Figure 2. Input data such as Latitude & Longitude for the desired location and Forecasting time is required for both Solar & Wind forecasting. Additionally, apart from the above inputs, for wind plant, hub height is required for calculating wind speed. Once these inputs are available, data is extracted from processed files stored in Cloud platform storage bucket. These files consist of Satellite Modeled Data at 0.25 deg grid size from 1999 onwards.

[0056] • Parameters: oGlobal Horizontal Irradiation (GHI) (2m height) oWind Speed at 10m height oWind Speed at 100m height

[0057] Said data is further downscaled from mesoscale to point location using statistical methods. Once the data is extracted & processed, the output data is ready for step2.

[0058] 2). Vertical extrapolation & extracting other Satellite Data

[0059] This step is particularly for Wind speed Forecasting.

[0060] Referring to Figure 3, from output of stepl, the data is required to be vertically extrapolated to hub height to create synthetic wind speed at hub height. This requires information such as wind surface pressure, wind speeds at different pressure levels. Data is extracted from processed files stored in Cloud platform storage bucket. These files consist of Satellite Modeled Data at 0.25 deg grid size from 1999 onwards for surface pressure & wind speed at different pressure levels. The data is further downscaled from mesoscale to point location using statistical methods. Once the data is downscaled, Empirical formula is applied to calculate height at each pressure levels using information from surface ground pressure as mentioned below:

[0061] The output from the above empirical formula provides height at different pressure levels from ground surface. Said information is further ingested in Power Law formula to calculate wind shear factor for extrapolating wind speed at hub height level using the empirical formula as mentioned below:

[0062] Wherein;

[0063] 3). Processing Extracted data using local Meteorological station data

[0064] (This step is particularly for Wind speed Forecasting.)

[0065] The wind is severely slowed down by obstacles and the roughness of the ground, hence there is a need to understand the local influence / ground measured data. This information requires to correct the output from step 3.

[0066] Referring to Figure 4, data is extracted from processed files stored in the storage means (110). Said files consist of Satellite Modeled Data at 0.25 deg grid size from the year 1999 onwards at 2 different levels. The data is further downscaled from mesoscale to point location using statistical methods. Data from local Meteorological station is extracted through API. Once both of the datapoints are extracted, factor is identified based on deviation in 2 different locations at 2 different elevations. This factor is used to recorrect the Wind speed.

[0067] 4). PRE- ANALYSIS

[0068] Data preparation before Machine Learning model preparation

[0069] (This step is for both for wind speed & solar GHI forecasting.) Referring to Figure 5, in this step the data is prepared for forecasting using GCM model forecasted data since 2017. A General Circulation Models (GCM) is a climate model that employs a mathematical model of the general circulation of a planetary atmosphere or ocean. It uses the Navier-Stokes equations on a rotating sphere with thermodynamic terms for various energy sources (radiation, latent heat).

[0070] Data is extracted from processed files stored in the storage means (1 10). Said files consist of GCM Modeled Data at 1 deg grid size. The data is further downscaled from mesoscale to point location using statistical methods. In the next step, specifically for wind, the data is recalibrated to ensure optimal distribution of wind speed at each frequency level using machine learning model. Further, using feature engineering techniques data is prepared based on seasonality & other factors and the datapoints are split in training & forecast datasets. ). GENERATING FORECAST

[0071] (This step is for both for wind speed & solar GHI forecasting.)

[0072] Referring to Figure 6, in this step Machine learning model is developed to train the model & forecast. The model trains on training dataset with auto-tunning the hyperparameters. Using the best parameters, forecast is generated.

[0073] • Solar - GHI (lYear & 25 Years).

[0074] •Wind - Wind Speed & Direction (lYear & 25 Years). 6). POST ANALYSIS

[0075] (This step is for wind speed forecast.)

[0076] Referring to Figure 7, in this step post forecasting wind speed, Machine learning model is used to fine tune the output and ensuring optimized frequency distribution bins.

[0077] Further, the efficiency of the present system can be verified in terms of accuracy and precision with the help of Test data as mentioned below:

[0078] TEST DATA:

[0079] WIND SPEED *

[0080] Table 1

[0081] Referring to the Table 1, it is understood that the wind speed was estimated using present system at two different sites i.e., Gujarat and Madhya Pradesh. The accuracy using the earlier conventional methods was only 88-90% at Gujarat and 87-91% at Madhya Pradesh. However, the accuracy of the present invention is significantly greater at 96%. 2.

[0082] Table 2

[0083] With reference to Table 2, it is observed that the accuracy in the estimation of solar irradiation using present invention was more than 97% as compared to the accuracy of the earlier used conventional methods. ADVANTAGES OF INVENTION:

[0084] The present invention is a paradigm shift in how solar and wind resources are estimated.

[0085] With inclusion of Climate change impact with other socioeconomic factors, the present invention accurately forecast the resource availability at any location in the world.

[0086] The present invention forecasts wind and solar resources for next 25 years with 96% and 98% annual accuracy which is 6% and 4% higher respectively than the earlier method.

[0087] Even without local measured data (wind mast with anemometer and pyranometer), the present invention is able to accurately forecast solar and wind resource. Hence, it does not require site measured historical data to forecast solar and wind resources for that site.

[0088] The present invention is capable to forecast wind and solar resources at any point location in the world.

Claims

Claims1. A System for accurately determining Long-Term Forecast comprises:•An anemometer (100)•A controlling means 1 (102)•A weather monitoring station (104)•A controlling means 2 (106)•A communicating means (108)•A storage means (110)Characterized in that;•A processing means (112)Wherein;Said anemometer (100) installed on wind turbine at hub height and connected to said controlling means 1 (102) is configured to count the number of rotations, or turns of the turbine; said Controlling means 1 (102) is configured to relay the data received from said anemometer (100) to a Controlling means 2 (106);Said weather monitoring station (104) connected to said controlling means 2 (106) is configured to measure parameters such as but limiting to global solar horizontal irradiation (GHI), inclined irradiation, wind speed, rain fall, and temperature;Said controlling means 2 (106) is configured is configured to relay the data from said weather monitoring means (104) to said communicating means (108) and then to said storage means (110);Said storage means (110) not limiting to a cloud server is configured to form an input database for said processing means (112);Said processing means (112) is configured to receive the input data stored in said storage means (110) and further analyse said data to provide an accurate long-term weather forecast. he System for accurately determining Long-Term Forecast as claimed in claim 1 wherein, the method for the working of said system comprises steps:• Satellite Data Extraction• Vertical extrapolation & extracting other Satellite Data• Processing Extracted data using local Meteorological station data• Pre-analysis• Generating Forecast• Post-analysis he System for accurately determining Long-Term Forecast as claimed in claims 1 and 2 wherein; said controlling means 1 (102) is preferably but not limiting to a Programmable Logic Controller (PLC) system.he System for accurately determining Long-Term Forecast as claimed in claims 1 to 3 wherein; said controlling means 2 (106) is preferably but not limiting to a Supervisory Control and Data Acquisition (SCADA). he System for accurately determining Long-Term Forecast as claimed in claims 1 to 4 wherein; said communicating means (108) is preferably but not limiting to an Open Platform Communications (OPC) Server. he System for accurately determining Long-Term Forecast as claimed in claims 1 to 5 wherein; said storage means (1 10) is preferably but not limited to a Cloud server.

Images