WO2013102932A2 - System and method facilitating forecasting, optimization and visualization of energy data for an industry - Google Patents

System and method facilitating forecasting, optimization and visualization of energy data for an industry Download PDF

Info

Publication number
WO2013102932A2
WO2013102932A2 PCT/IN2012/000845 IN2012000845W WO2013102932A2 WO 2013102932 A2 WO2013102932 A2 WO 2013102932A2 IN 2012000845 W IN2012000845 W IN 2012000845W WO 2013102932 A2 WO2013102932 A2 WO 2013102932A2
Authority
WO
WIPO (PCT)
Prior art keywords
data
forecasting
energy
optimization
forecast
Prior art date
Application number
PCT/IN2012/000845
Other languages
French (fr)
Other versions
WO2013102932A3 (en
Inventor
Gopi SUBRAMANIAM
Kaushal Kishore
Praveen Kumar SRINIVASAN
Vijaya Saradhi GANNAVARAM
Roopesh Ranjan
Awadhesh Kumar
Rishikes SAPRE
Srikanth Krishna RAJAGOPALAN
Original Assignee
Mzaya Private Limited
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mzaya Private Limited filed Critical Mzaya Private Limited
Publication of WO2013102932A2 publication Critical patent/WO2013102932A2/en
Publication of WO2013102932A3 publication Critical patent/WO2013102932A3/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes

Definitions

  • the present invention in general relates to a system and method for processing energy data for one or more utility purpose. More particularly, the invention relates to a system and j; method to provide forecasting, optimization and visualization of one or more parameters of energy data.
  • the power industry value chain comprises Power Generators, Distributors and Transmitters, each having their own conditions that change dynamically to affect the available power to end consumer at the same time the price at which this power is made available.
  • the availability of sufficient power to consumers whether industrial, agricultural or on household is important for trade and industry growth and human development.
  • the drawl schedule on the demand side is continually changing because of various observed and unobserved factors like weather condition, events and power conservation or recession etc. While there are macro factors on the supply side which affect the available capacity such as availability of power from different sources, generator condition and reliability associated with the supply. Hence the power which can potentially be made available is fixed except for sources like wind or solar power where the available power is a function of weather parameters.
  • the distributors face the challenge of anticipating the likely demand schedule and accordingly schedule power with generators.
  • the generators have to ensure their ability to ramp up or ramp down units in order to be able to meet the desired load shape almost real time given that energy as a commodity which cannot be stored.
  • price of power has become focus of all transactions. For example, after the introduction of Availability Based Tariff and power exchanges, forecasting of price of power has become important. Price forecasting has become the important aspect of power market management. Thus accurate price forecast data is need of the power generators, distributors and market players for optimization of power production, distribution and bidding strategy. There is necessity of this information in optimal scheduling of hydro energy production or thermal energy production with minimum constraints.
  • An intelligent system of control is therefore required to ensure that the positions on the demand and supply side are appropriately mapped in real time. It also needs to be ensured that the available supply sources are distributed in such a way that the existing transmission does not become a constrain where everyone starts drawing from a particular zone just because the tariffs were more lucrative but then could not deliver to the end consumer due to transmission congestions.
  • One of the prior art discloses a power market management where a plurality of process streams having inter related process streams are established.
  • these methods are intended to be used by an electric power dispatch centre or an exchange where price, day- ahead schedule, etc is optimized.
  • This patent pertains to solving the problem of an independent system operator of managing the power trade/managing the share of different utilities.
  • the key problem is to know the status of each generator, know the share of each utility in every generator, the position of each utility based on the availability, managing the transmission corridor and managing the power trade for utilities as independent exchange.
  • it provides solution to individual dispatch entity.
  • the decision support system here is intended to be used by generators to determine generation of electricity by integrating a wide variety of disparate data into a comprehensive information set that can be used to determine when, where, and how to generate electrical power.
  • this invention is solving the problem from the generation side with different challenges.
  • One of the prior art discloses a demand forecasting system for the day-ahead using historical demand data, weather forecast for the next day and calculating the impact of day-of-the-week. This is using a time-series and regression based approach. This invention uses a similarity based approach for calculating weather impact by finding a day with similar weather in the past.
  • a computer implemented risk-management system schedules the generating units of an electric utility while taking into consideration power trading with other utilities and the stochastic load on the utility system.
  • the system provides the user multiple load forecasts and allows the user to vary the fuel price between different scenarios and different periods of planning horizon.
  • this system focuses on optimizing the price of fuel from generation side only. Thus this solution has limited approach.
  • the present invention discloses an intelligent system to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data.
  • the system comprises of a data gathering means configured to gather the energy data from one or more components distributed throughout the network, a data sorting module configured to categorize said energy data and transmit it to one or more sub-systems for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement and a network of subsystems, to receive categorized data for generating plurality of end results.
  • the network further comprises of a forecasting sub system configured to process the categorized data and related constraints for the purpose of one or more type of forecasting for a particular parameter by applying one or more type of forecasting operations and an energy optimization sub system configured to process the categorized data and related constraints by means of its data processing components and generate one or more optimal solution for one or more parameters by using a dynamic optimization technique.
  • the system further comprises of a communication channel configured to integrate the sub systems distributed throughout the network and enable the communication between them, the communication channel further enables an on-demand exchange of energy data and the end results amongst these sub systems and an output generation unit configured to communicate with one or more sub systems and captures the end results for further post-processing.
  • the system further comprises of a data visualization module configured to further post-process the end result of one or more sub systems by using a suitable visualization methodology in accordance with the particular end result, such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry.
  • a data visualization module configured to further post-process the end result of one or more sub systems by using a suitable visualization methodology in accordance with the particular end result, such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry.
  • the present invention also discloses an intelligent method to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data.
  • the method comprises of steps of gathering energy data from one or more components distributed throughout the network, categorizing said energy data and transmit it to one or more subsystems for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement and generating plurality of end results by receiving categorized data.
  • the generation further comprises of processing the categorized data and related constraints for the purpose of one or more type of forecasting for a particular parameter by applying one or more type of forecasting operations and for energy optimization by means of one or more data processing components and generating one or more optimal solution for one or more parameters by using a dynamic optimization technique and processing the categorized data and related constraints.
  • the method further comprises of integrating the sub systems distributed throughout the network and enabling the communication between them, the integration further enables an on-demand exchange of energy data and the end results amongst these sub systems, capturing the end results of one or more sub systems for further post processing and post-processing the end results of sub systems by using a suitable visualization methodology in accordance with the particular end result, such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry.
  • the present invention also discloses a forecasting system to determine one or more types of forecasting results associated with energy for one or more sector distributing said energy.
  • the system comprises of an input capturing device configured to capture data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module, a data filtration module configured to pre-process the data by using one or more data filtration techniques to remove an unwanted data and a processor configured to determine by way of its embedded algorithms one or more types of forecast results for a pre-defined time period.
  • the processor further comprises of one or more sub- processing modules configured to apply one or more forecasting operations, such that the operations are applied with respect to the data thus captured and a rule based engine configured to perform one or more post processing steps over the forecast results by using predefined rules to further correct said forecasted results, the system further comprises of an output generation module configured to generate one or more types of corrected forecast results at one or more level associated with the energy distribution sector with respect to each source system.
  • the present invention also provides a forecasting method to determine one or more types of forecasting results associated with energy for one or more sector distributing said energy.
  • the method comprises of steps of capturing data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module, pre-processing the data by using one or more data filtration techniques to remove an unwanted data and processing the data to determine by way of embedded algorithms one or more types of forecast results for a pre-defined time period.
  • the processing further comprises of applying one or more forecasting operations, such that the operations are applied with respect to the data thus captured.
  • the method further comprise of steps of post-processing the forecast results by using predefined rules to further correct said forecasted results and generating one or more types of corrected forecast results at one or more level associated with the energy distribution sector with respect to each source system.
  • the present invention also discloses a system to facilitate energy optimization with respect to one or more constraints for an industry type for a utility purpose.
  • the system comprises of an integration module configured to integrate one or more source systems distributed across a network, a data capturing device configured to pull data from the source systems thus integrated throughout the network, the data capturing device further comprises of a constraint fetching module configured to take into account the constraints associated with the data thus pulled for a particular industry type and a processing engine configured to process the data in combination with the constraints to determine an optimized solution with respect to said constraint.
  • the processing engine further comprises of an optimization module configured to form an objective function of said constraint by running a space reducing robust mechanism for reducing errors while determining the optimal solution.
  • the system further comprises of an output generation module configured to generate the optimal solution with respect to the data thus captured, such that the optimal solution are used for one or more utility purpose.
  • the present invention also discloses a method to facilitate energy optimization with respect to one or more constraints for an industry type for a utility purpose.
  • the method comprises of steps of integrating one or more source systems distributed across a network, pulling data from the source systems thus integrated throughout the network and taking into account the constraints associated with the data thus pulled for a particular industry type.
  • the method further comprises of steps of processing the data in combination with the constraints to determine an optimized solution with respect to said constraint.
  • the processing further comprises of forming an objective function of said constraint by running a space robust mechanism for reducing errors while determining the optimal solution and generating an optimal solution with respect to the data thus captured, such that the optimal solution is used for one or more utility purpose.
  • FIG. 1 illustrates the architecture of intelligent system in accordance with an embodiment of the invention.
  • Figure 2 illustrates a flow chart towards forecasting, optimization and visualization energy data in accordance with an embodiment of the invention.
  • Figure 3 illustrates the architecture of forecasting system in accordance with an alternate embodiment of the invention.
  • Figure 4 illustrates the architecture of energy optimization system in accordance with an alternate embodiment of the invention.
  • Figure 5 illustrates different type of forecasting and optimization by exchanging data in accordance with an exemplary embodiment of the invention.
  • Figure 6 illustrates features of optimization in accordance with an exemplary embodiment of the invention.
  • Figure 7 illustrates the flow of data in a network in accordance with an exemplary embodiment of the invention.
  • modules may include self-contained component in a hardware circuit comprising of logical gate, semiconductor device, integrated circuits or any other d iscrete component.
  • the module may also be a part of any software program executed by any hardware entity for example processor.
  • the implementation of module as a software program may include a set of logical instructions to be executed by the processor or any other hardware entity. Further a module may be incorporated with the set of instructions or a program by means of an interface.
  • the present invention provides a system and method for processing energy data and generating one or more end results in a network. These results may relate to forecasting, optimization and visualization of one or more parameters associated with data. While processing the data the system also takes into one or more constraints and generates results accordingly.
  • the data is gathered from plurality of sources or components distributed throughout the network. The data is pre-processed and the end results are post processed in order to further provide visualization of the data which may provide suggestive measures be used by one or more industry.
  • the present invention comprises of an intelligent system (100) to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network. End results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data.
  • the intelligent system (100) further comprises of a data gathering means (102) configured to gather the energy data from one or more components distributed throughout the network and a data sorting module (104) configured to categorize said energy data and transmit it to one or more sub-system (106) for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement.
  • the intelligent system (100) further comprises of a network of subsystems (106) to receive categorized data for generating plurality of end results.
  • the network of subsystems (106) further comprises of an energy optimization sub system (108) and a forecasting sub system (110) configured to process the categorized data and related constraints.
  • the intelligent system (100) further comprises of a communication channel (112) configured to integrate the sub systems distributed throughout the network and enables the communication between them, an output generation module (1 14) a data visualization module (1 16).
  • the integrated system (100) providing forecasting, optimization and visualization of one or more parameters associated with energy data may include but is not limited to electrical energy, water supply.
  • the end results of forecasting, optimization and visualization depicts suggestive measures on considering one or more constraints associated with said data
  • constraints may include but is not limited to power availability, unit price of power in exchange and grid, price and volume constraint for each source, grid stability related constraints, technical constraints relate to running the power plants, reliability requirement of each node in the network, to power surplus, power short fall estimation, contract.
  • the data gathering means (102) gathers the energy data from one or more components distributed throughout the network (step 202 of figure 2). These components may also be referred to as source components or source systems.
  • the energy (electric energy) data which is thus gathered is common and this data should be categorized so that it may be used and fed for the purpose of forecasting, optimization and visualization.
  • the data sorting module ( 104) performs this task of categorizing this data and then transmitting it to the respective subcomponent (steo 204 of figure 2).
  • This data gathering module (102) acts as a source system for all the sub components processing the data.
  • the forecasting system (300) (also referred as forecasting sub system) further comprises of an input capturing device (302) configured to capture data from one or more source systems, a data filtration module (304) and a processor (306).the processor further comprises of one or more sub-processing modules (308) configured to apply one or more forecasting operations;
  • the forecasting system (300) further comprises of a rule based engine (310) configured to perform one or more post processing steps over the forecast results, an output generation module (312) configured to generate one or more types of corrected forecast results.
  • the forecasting system (300) implements a method to determine one or more types of forecasting results associated with energy for one or more sectors distributing said energy (step 208 of figure 2).
  • the energy further includes electrical energy.
  • One or more types of forecast results further comprises of results for Electricity demand , forecasting, UI (Unscheduled Interchange) Forecasting and IEX (Exchange Price) Forecasting.
  • the forecasting system (300) further comprises of an input capturing device (302) configured to capture data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module.
  • the data captured by an input capturing device (302) further comprises of data related to historical electricity/energy consumption and distribution data, weather data, grid-wise data, population and past demand in order to determine electricity demand forecasting, UI(unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
  • the data captured by an input capturing device (302) in order to determine electricity demand forecasting may include but is not limited to Historical Electricity Demand, Historical Weather, Forecasted Weather, Weather Variation Range, Holiday List, Population Growth Rate, Energy Consumption Growth Rate, GDP Growth Rate, Industrial Output Growth, Agricultural Growth Rates, etc.
  • Historical load data includes details on power cuts to calculate Unrestricted Peak Demand (UPD). If this data is available at a segment or feeder level, a breakdown of demand at these levels can be obtained.
  • Weather Data includes Temperature, Humidity, Rainfall, Wind Speed and Cloud cover. Cluster/ location details at a desirable level, Calendar Data including most of the calendar effect details which are commonly available, if there are area specific events/calendar effects that also included, Events details and list and details of power sources (PPAs, Own generation details etc.).
  • the data captured by an input capturing device (302) in order to determine UI (Unscheduled Interchange) forecasting includes Grid-wise Historical drawl data aggregated over all distributors in the grid, Grid-wise schedule aggregated over all distributors, Grid wise generation, Historical UI values for the grid, Historical weather for representative locations in the grid, weather forecast for representative locations in the grid, Holiday list.
  • the data captured by an input capturing device (302) in order to determine IEX (Exchange price) forecasting includes Historical Exchange Price, Historical weather for representative location, condition of generating units that feeds in power to this region, the demand pattern (fluctuations) for every beneficiary in this region, weather forecast for representative locations in the grid, Holiday list.
  • the data filtration module (304) of forecasting system (300) is configured to pre-process the data by using one or more data filtration techniques to remove an unwanted data.
  • the data filtration module (304) of forecasting system (300) is configured to pre-process the data by using one or more data filtration techniques to remove an unwanted data.
  • There can be errors in the captured data due to variety of reasons such as bad data capture practices, distribution or transmission failure etc. These aberrations in the data may result in an incorrect relation between input and output. So in order to improve the accuracy of the model and a better understanding of the relationship, pre-processing step is carried out to clean the captured data.
  • Data preprocessing step is carried out by using one or more data filtration techniques.
  • the data filtration techniques used may include use of an adaptive median filter to filter the data.
  • An adaptive median filter applied for each time block to mark unusually high/low Demand/Weather/UI/Frequency values. If the data is suspected, it is either discarded or a suitable replacement is found based on historical data/basic forecast. Finally the filtered data is used further in forecasting.
  • the processor (306) is configured to determine by way of its embedded algorithms one or more types of forecast results for a pre-defined time period.
  • the forecast results for predefined time period ranges from 15 minutes to 15 years.
  • the processor further comprises of one or more sub-processing modules (108) configured to apply one or more forecasting operations, such that the operations are applied with respect to the data thus captured.
  • the forecasting operations further comprises of a regression based approach, additive splines, parameter shrinkage, grid-level demand/schedule/generation and generic algorithms or a combination thereof in order to determine electricity demand forecasting, UI(unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
  • the forecasting operation steps referring to Electricity demand forecasting includes but not limited to following steps.
  • Regression base approach is used to relate the future demand to- historical demand, weather, events, planned schedules etc.
  • the regression tries to identify relationships between future demand and these parameters. These relationships are then used to project future demand.
  • Additive Splines are used to capture non linearity with weather and Demand. Parameter shrinkage is done to avoid over-fitting. Separate model is built for each time block. Adaptive mean error adjustment is done to capture local effects. Holidays captured through a factor based model
  • Probabilistic density forecasting is used to capture forecasting uncertainty. In some cases, an ensemble of multiple methods was used to get a better result.
  • Regression based approaches and neural network is used with different inputs to explore different kinds of relationship between inputs and the demand. Further an ensemble method is used to combine these relationships into one that gives a better prediction.
  • the forecasting operation steps referring to UI (unscheduled Interchange) forecasting includes but not limited to Regression based approach and use of additive splines.
  • the forecasting operation steps referring to IEX (Exchange price) forecasting includes but not limited to: Applying Regression based approach using historical IEX data.
  • Current generation Injection schedule is captured in the model.
  • Current grid shortage is captured in the model.
  • Buy and sell bidding is captured by using "sigma" factors (Square root of average deviation A 2 between forecast and actual). This is further used to optimize the range that determines the sale and buy bids. Bid clearance and buy-sell price gap is simultaneously optimized. Alternatively, in some cases, genetic algorithms are also used for the same.
  • the forecasting operations are performed by the sub processing modules such that the operations are applied with respect to the data thus captured and the required forecasting result type such as electricity demand forecasting, UI (unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
  • UI unscheduled Interchange
  • IEX Exchange price
  • the rule based engine (310) is configured to perform one or more post processing steps over the forecast results by using predefined rules to further correct said forecasted results.
  • the rule based engine is configured to perform one or more post processing steps further comprises of a use of Kalman filter in order to sort the errors in historical data.
  • the data post processing steps includes making use of specific rules developed for each utility for example, Correction for Agricultural Load Change, Correction for planned events etc.
  • Further Post-processing techniques also includes use of Kalman Filter in order to do an error correction and thus eliminate un-wanted noise in forecasting. This works as a local model to correct local deviations between forecasted and actual parameters.
  • the rules are used to override or correct the system forecasts. There are two kinds of corrections. In the first way, it is done by the system itself where it looks at the historical errors and corrects itself to get closer to the actual. In the second way, it is done manually, where the user can change the forecast based on their understanding of the market.
  • the system takes the inputs from the users in terms of whether the demand will go up or down in a block or set of contiguous blocks and ensures that the demand pattern is maintained while increasing or reducing the demand.
  • the output generation module (312) of forecasting system (300) is configured to generate one or more types of corrected forecast results at one or more levels associated with the energy distribution sector with respect to each source system.
  • the one or more types of corrected forecast results further comprises of a point forecast, a range or a probability distribution or a combination thereof with respect to the electricity demand forecast, UI/Frequency Forecast at 15 minute interval with respect to IEX (Exchange Price) Forecasting and Lower and upper limit for exchange Price at 15 minute interval for the specified region with respect to IEX (Exchange Price) Forecasting.
  • One or more levels associated with the energy distribution sector further comprises of an aggregate distributor level, station level, sub-station level, feeder level or a combination thereof.
  • the corrected forecast results generated with respect to electricity demand forecasting includes electricity Demand Forecast for the distributor at multiple horizons. More particularly it is for an hour ahead, day-ahead, day- week ahead, months ahead, year ahead, and 3-20 years ahead.
  • the output can be in the form of a point forecast, a range or a probability distribution.
  • the forecasts can be generated at an aggregate distributor level, station level, sub-station level, feeder level etc. This forecast is used in Power Procurement to meet demand as well as having some flexibility to do power trading within the limits specified by the regulator.
  • electricity demand forecast is generated as short term forecast and medium term and long term demand forecast.
  • the short-term demand forecast module generates forecast for individual blocks (96 blocks of 15 minutes interval) at the following level of granularity including, day ahead, week ahead, max, min and volatility. It can be extended to generate this forecast at various level of aggregation across dimension based on geography and usage. For example, the user might want to generate area or cluster wise load forecast for industrial, domestic, commercials, government offices, railways etc. This depends on the availability of data at the desired level.
  • medium term demand forecast can be generated at hourly or peak and off-peak on a daily level.
  • Daily Forecast of Peak Demand, Base Demand for coming 1 year is predicted.
  • the forecast is divided into multiple blocks of a day such as Early morning demand, Late morning demand, Afternoon demand and Late evening demand.
  • a range of demand for these blocks at a pre-specified confidence level is predicted which includes Most likely demand, As-high-as demand and As-low-as demand.
  • medium term demand forecast includes generating year ahead demand forecast for 15 minutes interval or 3 to 5 years demand forecast for 1 hour interval.
  • long term demand forecast includes generating demand forecast for 25 years at monthly interval.
  • the corrected forecast results generated with respect to UI (unscheduled Interchange) forecasting includes Unscheduled Interchange or Grid Frequency Forecast at 15 minute interval to several days ahead.
  • the forecast result can be used by the distributor to anticipate likely values of the grid frequency during the day leading to optimal planning, procurement, shutting down/ramping up of captive generation, etc.
  • the corrected forecast results generated with respect to IEX (Exchange Price) forecasting includes lower and upper limit for exchange Price at 15 minute interval for the specified region. Further in exchange price forecasting, forecasting horizon can be from 15 minutes to several days ahead.
  • the exchange price forecasting module predicts the market clearing price for multiple regions. The range in which the market clearing price will fall is predicted. This helps the user in deciding the bidding price for every block in a day. This forecast can be used by the distributor to anticipate likely values of the exchange price during the day leading to optimal planning, procurement, shutting down/ramping up of captive generation, etc.
  • the demand forecast result decomposition is carried out by area, region or cluster.
  • the forecast result can be further decomposed by type of customers wherein type of customers can be further categorized as migrated customers, own customers and open access customers. Further, custom grouping can be specified for example by per capita, load factor, density and diversity factor.
  • features of the forecast result more particularly electricity demand forecast comprises of capturing non-linear relationship with weather parameters, Modelling load variation during a day, Modelling weekday, weekend variation, Capturing Holiday effect, applying adaptive correction to learn from past forecasting errors, providing facility to incorporate domain expertise input into forecast output, carrying out risk or variation quantification of the forecast errors in different situations, carrying out optimistic and Pessimistic Scenario Analysis for multiple weather conditions, taking into account Economic, Demographic or Geographic parameters like GDP Growth, Energy consumption per capita growth, population under electricity growth, industrial growth for long term forecast.
  • features of the forecast result more particularly UI (Unscheduled Interchange) Forecast comprises of capturing day of the week effect, capturing time of the day effect, capturing holiday effect, capturing non-linearity of weather parameters and analyzing risk with respect to any production unit failing or abnormal increase in demand.
  • features of the forecast result more particularly IEX (Exchange Price) Forecast comprises of capturing day of the week effect, capturing time of the day effect, captures holiday effect, capturing non-linearity of weather parameters and analyzing risk of price and volume.
  • features of the forecasting system includes adaptable to new data and parameters, flexible to accommodate features including event database to capture special conditions, scale up and scale down forecast at overall and individual block levels, forecasting at different granularity, region or cluster level forecast, customer segment level forecast and simple to use and interactive modelling platform.
  • the forecasting system (300) enables the user to use his domain knowledge or any region or date specific additional information to change the system generated forecast.
  • the revised forecast is also stored in the system and used for any future analysis which becomes very useful in the case of special events that are not declared in advance.
  • the system (100) now further comprises of a sub system to facilitate energy optimization (also referred to as energy optimization or simply optimization system) which is present in the network of sub-systems and is integrated with all the other sub-systems (step 210 of figure 2).
  • energy optimization also referred to as energy optimization or simply optimization system
  • the energy optimization system (400) for energy optimization with respect to one or more constraints comprises of an integration module (402) configured to integrate one or more source systems distributed across a network.
  • the energy optimization system (400) further comprises a data capturing device (404) to pull data from the source systems throughout the network.
  • the data capturing device (404) further comprises a constraint fetching module (506) configured to consider the constraints associated with the data pulled from the source systems.
  • the optimization system (400) further comprises of a processing engine (408) configured to process the data in combination with the constraints to determine an optimization solution with respect to said constraint.
  • the processing engine (408) further comprises an optimization module (410) configured to form an objective function of said constraint by running a space robust mechanism for reducing errors while determining the optimal solution.
  • the energy optimization system (400) comprises of an output generation module (412) configured to generate the optimal solution with respect to the data thus captured.
  • the system (400) to facilitate energy optimization for an industry type may include but is not limited to electrical energy, water supply.
  • the energy optimization system (400) comprises of said integration module (402) configured to integrate said one or more source systems distributed across network.
  • the source system comprises of one or more forecasting systems, PPA, Hydel generation, Thermal generation, Gas generation, Bilateral, Exchange, UI.
  • the data capturing module (404) is configured to pull the data from the source systems thus integrated throughout the network.
  • the data pulled from the source systems may include but not limited to the data related to price and availability of each of power sources, price of deviating from the agreed volume, reliability of each power source, corridor limit and transmission constraints, demand forecast, UI forecast, exchange price forecast, technical specifications on each type of power source.
  • the data capturing module (404) pulls the data from the source systems in real-time.
  • the data capturing device (404) further comprises of the constraint fetching module (406) configured to take into account the constraints associated with the data pulled for particular industry type.
  • the constraints associated with the data may include but not limited to power availability, unit price of power in exchange and grid, price and volume constraint for each source, grid stability related constraints, technical constraints relate to running the power plants, reliability requirement of each node in the network. Further, the constraints associated with the data may include but not limited to power surplus, power short fall estimation, contract.
  • the energy optimization system (400) comprises of the processing engine (408).
  • the processing engine (408) is configured to process the data in combination with the constraints to determine optimized solution (not shown) with respect to the constraint.
  • the processing engine (408) further comprises the optimization module (410) configured to form objective function of the constraint by running said space reducing robust mechanism (which is a dynamic optimization technique) for reducing errors while determining the optimal solution.
  • the constraint reducing robust mechanism (not shown) further comprises of a mathematical programming approach, a heuristic approach or a combination thereof.
  • the output generation module (412) generates the optimal solution with respect to the data thus captured, such that the optimal solution is used for one or more utility purpose.
  • the optimal solution further comprises of a short term energy portfolio optimization and a medium term energy portfolio optimization.
  • the energy optimization system provides following end results:
  • Short term Portfolio Optimization The energy optimization system provides the solution to the short term portfolio optimization in a unified unit commitment problem as well as financial decisions relating to short term bilateral, PPP, Day ahead forward and spot market. Price as well as volumetric risk is considered for risk management.
  • the system integrates an Industrial book (details of the generation units owned and managed by the utility) along with a Trade book (details of all external sources of power with details of terms and price associated with each agreement) for the optimal decision making process.
  • the utility can fulfil that demand by a combination of supply from different sources (internal as well as external detailed in the Industry and Trade Book).
  • sources are associated with an entitlement of power for every time- block and cost for deviating from that entitlement.
  • the salient features of the short-term optimization module are as following:
  • step 502 It pulls in the demand forecasting (step 502 in figure 5) for the next day from a demand forecasting module (forecasting system).
  • a demand forecasting module forecasting system.
  • Several forecasts may be generated for a day by providing different inputs or by revising the forecast generated by the system (step 504 of figure 5).
  • the energy optimization further provides a flexibility to the user to change any of these data before running the dynamic optimization technique. For example, if a particular plant is shut down for maintenance, the user can remove it from the optimization. Similar changes can be done for trade book, if required.
  • the system automatically connects to SLDCs and fetches the real-time running capacity of each unit of the plants in the industry book. Based on this real-time data, it predicts the running capacity of the units for the next day. On top of this, the user can change the capacity at a block level based on any additional information.
  • the exchange forecasting can also provide a range rather than a point forecast. This will help the utiltiy in deciding on the bidding strategy.
  • the key objective is to provide a decision that minimizes the overall cost of power procurement.
  • the decisions include how much power should be taken from each source, how to do a trade-off between exchange and UI, how to schedule the generation plants etc.
  • the energy optimization system also provides the total cost for this portfolio and the risk measures around this decision.
  • the inputs to the optimization system include several forecasts (demand forecast, price forecast, availability forecast etc.) and hence the quality of the solution is dependent on the quality of these inputs. However, it is obvious that these inputs will have uncertainty associated around it. If the optimization is done using these estimates as crisp inputs without understanding the uncertainty, the optimization solution may not remain optimal (or even feasible) if these inputs change. Therefore, there is a need of a solution that is more robust to these changes in the inputs.
  • the energy optimization system uses state-of-the-art Robust Optimization approach to take care of variability in the inputs and uncertainty in the environment to provide a more robust decision.
  • the system uses the following approach to introduce robustness in the system.
  • the system has a sensitivity analysis to enable users to do what-if analysis.
  • the system generates multiple estimates to capture different scenarios. Based on these multiple estimates, a most-likely and two extreme estimates are generated. 2.
  • the optimization problem is converted into a multiple objective optimization problem. The first objective is to optimize the most-likely case. Whereas the other two objectives intend to minimize the loss in worst case and maximize the gain in the best case situation.
  • the energy optimization system (400) further integrates with external systems for workflow to enable communication of decisions between teams (back-office, middle-office and trading teams) and getting feedback.
  • the energy optimization system is automated enough to minimize human error, at the same time, can incorporate feedback provided by experts. This system takes care of technical constraints (swing available on a PPA, UI limit or idling and switching constraints on power plants etc.) as well as business constraints (high reliability for special areas, distribution of power between urban and rural areas etc.).
  • the system helps the user to make decisions on buying or selling of power in the medium term (upto 1 year ahead). It provides recommendation on buying/ selling RTC power and block-wise power. Furthermore, since we are considering a longer time horizon the system does not provide one recommendation for each period. It provides multiple recommendations based on the confidence level.
  • the key features of the medium and long term optimization modules are as follows: a. Takes the inputs from the forecasting system of the intelligent system. b. Takes inputs from trade book and Industry book regarding existing agreements and available power from other sources. c. Calculate the Position Gap at different levels (RTC, Block-wise) d. Find the optimal mix of trade opportunities at different sources
  • the optimum solution is implemented by two methods- mathematical programming approach by way of combining integer linear programming problem and heuristic method in order to solve the problem in multiple steps.
  • the mathematical programming approach is explained in way of an example mentioned below. All these steps are performed by the processor thus present in the energy optimization system:
  • the integration module configured to integrate said one or more source systems comprising PPA, Hydel generation, Thermal generation, Gas generation, Bilateral, Exchange, UI.
  • the constraint fetching module considering the constraints associated with the data pulled the constraints associated with the data may be presented as follows:
  • drawl from ui is not more than x% of drawl from PPA, EXCHANGE & BILATERAL. This is a regulatory constraint and the maximum amount that one can overdraw from the grid at different frequency.
  • ppa_drawl ⁇ ppa drawl Limit (entitlement)(zero deviation allowed above it, but can plan in advance to draw less).
  • ppa_drawl_limit under_drawl_from_ppa + ppa drawl (variable from 4th constraint) 6.
  • Abrupt change constraint Powerplant generation cannot be changed very often, especially for thermal units. Once a change in generation is made, it needs to be the same for atleast 4 hours. Once it is shut down, it needs to be down for a day. For a gas or liquid plant, these changes can be made more frequently.
  • Availability of power plant there can be two kinds of hydel plants (reservoir based and run of the river based). In a reservoir based plant, the output can be controlled to a maximum of level of water, max capacity of the unit. In case of run-of-the-river unit, the running capacity will be in the same vicinity for consecutive days.
  • the processing engine processes the data in combination with the constraints to determine optimized solution with respect to the constraint.
  • the processing engine further comprising the optimization module forms objective function of the constraint by running a space reducing robust mechanism (not shown) for reducing errors while determining the optimal solution.
  • the objective function for exemplary purpose considered are- Minimize (variable_cost_ppa * ppa_drawl + fixed_cost_ppa*under_drawl_from_ppa + cost_hydel*hydel_drawl + cost_non_hydel * non_hydel_drawl + cost_exchange * exchange drawl + cost ui * ui_drawl ) - penalty * power_cut
  • the data is organized as a table with source name, source type, minimal enforced drawl as per the agreement, price, overall power available from it, power available for optimization and recommended volume which needs to be filled in.
  • the table is sorted by price in ascending order.
  • step 5 For sources with abrupt change constraint, step 5 is done so as to ensure that the constraint doesn't get violated. In this case, steps 3,4,5 are repeated iteratively to ensure that maximum power gets, utilized from sources with lower prices and at the same time ensuring that the abrupt change constraint doesn't get violated. The iterative process can be stopped when the final cost between two iterations is lower than a pre-specified threshold.
  • the optimal solutions are used for utility purpose selected from a group of Prioritize across contracts, spot market and captive generation, Amount of volume to be exercised across contract, how much to buy or sell (depending on the supply Available) from day-ahead exchange and spot market, how much to generate if at all at any given point in time, given current status of various generation units and their marginal cost and the spot price, Prioritize Ramp-up & Ramp Down, which unit to shut down in case of excess capacity or a combination thereof.
  • the time unit is changed from 15 minute block to off-peak, on-peak to month.
  • the output is presented as volume of power to be procured at each time unit, buy or sale of power and bidprice for exchange, how to run internal generation units, how much power to surrender, how much buffer to keep, what is the risk associated with the solution.
  • the energy optimization system shares data from almost all the sub systems and modules present in the network of sub systems in order to generate end results.
  • the shared data further includes price forecast, available power, contracts, committed power, generation cost schedule, gap, node reliability availability, risk appetite, demand forecast etc.
  • the network of subsystem further comprises of modules like network planning module, position mapping module and power flow module.
  • the network planning module is configured to capture the availability of supply from different sources, Identifying the risk attached (reliability) to the expected supply of electricity from different sources and providing scenario based avai lability of electricity from different sources.
  • the position mapping module is configured to aggregate the demand and supply in the market to understand the position gap at different granularity and evaluating whether it is for the entire day or for different blocks of the day, whether it is at an aggregate level or at a sub-station/ feeder level and provides the position gap.
  • the power flow subsystem is further configured to provide optimal schedules for generation units over a 24 hour period which in turn minimizes operating cost of generation plants.
  • constraints included here are load balance, spinning reserve, down time limits, C02 emissions and ramp rate limits.
  • the data (or end results) from these three or more modules is again shared by energy optimization system for further processing and processing end results for the energy optimization system (as shown in step 508 of figure 5).
  • the intelligent system (100) now further comprises of the communication channel (1 12) configured to integrate the overall system with all the sub-systems (and modules) and external components (step 212 of figure 2).
  • This integration is performed by communicating with individual integration module which is coupled with each sub systems (forecasting and optimization). Because of this integration, these sub systems may exchange data with each other in case of any demand. For example, the energy optimization system (or sub system) is pulling out data from the forecasting system (sub system).
  • the system (100) comprises of the output generation unit (1 14) which is again transmitting the end results to further components for further post processing (step 214 of figure 2).
  • the data visualization module (1 16) further executes steps of post-processing the end results of sub systems (106) by using a suitable visualization methodology in accordance with the particular end result such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry (step 216 of figure 2).
  • the suggestive measures may be taken (depending on the infrastructure and data availability), at utility level, sub-station level, feeder level or even at a customer level.
  • the visualization further comprises of a color code determining a size of a position gap on a continuous color gap.
  • the data visualization module (1 16) further comprises of a report generation module configured to generate reports that enable the users to analyze accuracy of the forecast result at several granularities.
  • the system provides a very transparent way to check the accuracy of the forecast result on a particular day, over a week or over last 100 days.
  • the forecasting system also enables the user to separate errors due to weather forecasting error and demand forecasting errors. On a daily basis, the system compares weather forecast versus actual weather to enable this analysis.
  • MAPE mean Absolute percentage Error
  • Block-wise actual MAPE report as a heat map f Model tracking Report y
  • Weather forecasting accuracy report f Visual comparison of demand forecast and actual for a particular day.
  • f Other Customized Report can be frozen during the requirement analysis phase.
  • the intelligent system (100) also allows the users to separate errors due to weather forecasting error and demand forecasting errors. On a daily basis, the system compares weather forecast vs actual weather to enable this analysis y Aggregated view of demand and supply to understand the gap f
  • An enterprise wide snap shot of position gap at varying levels of granularity y PowerMatics powerful visualization capability helps drill down todesired levels
  • the intelligent system (100) comes bundled with state of the art powerful visualization capability which allows the user to analyze position gap visually where the color code determines the size of the position gap on a continuous color scheme.
  • the system (100) facilitates multidimensional visual charts which are easily comprehensible. The charts can be displayed on computer screens, large scale displays, iPads etc.
  • the intelligent method further comprises step of measuring risk to further determine one or more factor leading to a risk with respect to energy data.
  • step 506 the data from electricity transaction book, trade book and industry book after processing goes to the energy optimization system.
  • the intelligent system (100) transmits end results of all the sub systems throughout the network by using various communication sources like intranet, internet etc.
  • the transmitted data could be easily accessed or controlled over PC, tablets and request for such data could also be sent through SMS emails etc.
  • Various purposes in which the end results could be further utilized are for contract and invoice, demand forecast and trading.

Abstract

The present invention provides an intelligent system and method to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data. The energy data is gathered from one or more components distributed throughout the network and further categorized and transmitted to one or more sub-systems for further processing. A network of sub systems like forecasting system and optimization system processes the data by using forecasting operations and optimization techniques and generates end results. These end results are the post processed to generate suggestive measures for any gap thus found for one or more sector of an industry.

Description

SYSTEM AND METHOD FACILITATING FORECASTING, OPTIMIZATION AND VISUALIZATION OF ENERGY DATA FOR AN INDUSTRY
FIELD OF INVENTION
The present invention in general relates to a system and method for processing energy data for one or more utility purpose. More particularly, the invention relates to a system andj;method to provide forecasting, optimization and visualization of one or more parameters of energy data.
BACKGROUND OF THE INVENTION
With the economic growth of the country, it results in heavy demand of energy. Need of energy is from household consumption, industrial consumption to hospitals, commercial establishments, transportation, communication to entertainment. The need of energy is on large scale and in various forms. Whenever there is insufficient power supply, these sectors are badly affected by power cuts. Energy demand changes every hour and it depends on many factors. Although the energy demand can change anytime, energy infrastructure should be capable of meeting the demand always.
The power industry value chain comprises Power Generators, Distributors and Transmitters, each having their own conditions that change dynamically to affect the available power to end consumer at the same time the price at which this power is made available. The availability of sufficient power to consumers whether industrial, agricultural or on household is important for trade and industry growth and human development.
The drawl schedule on the demand side is continually changing because of various observed and unobserved factors like weather condition, events and power conservation or recession etc. While there are macro factors on the supply side which affect the available capacity such as availability of power from different sources, generator condition and reliability associated with the supply. Hence the power which can potentially be made available is fixed except for sources like wind or solar power where the available power is a function of weather parameters.
On the other hand, the distributors face the challenge of anticipating the likely demand schedule and accordingly schedule power with generators. Alternatively the generators have to ensure their ability to ramp up or ramp down units in order to be able to meet the desired load shape almost real time given that energy as a commodity which cannot be stored. Moreover, one also need to ensure that given the transmission capacity one do not want to get caught in a situation where though the power is available and could not be transmitted due to transmission capacity constrain or congestion in network. Hence, it becomes utmost important to distribute power procurement across units optimally.
With the reformation in power industry, price of power has become focus of all transactions. For example, after the introduction of Availability Based Tariff and power exchanges, forecasting of price of power has become important. Price forecasting has become the important aspect of power market management. Thus accurate price forecast data is need of the power generators, distributors and market players for optimization of power production, distribution and bidding strategy. There is necessity of this information in optimal scheduling of hydro energy production or thermal energy production with minimum constraints.
An intelligent system of control is therefore required to ensure that the positions on the demand and supply side are appropriately mapped in real time. It also needs to be ensured that the available supply sources are distributed in such a way that the existing transmission does not become a constrain where everyone starts drawing from a particular zone just because the tariffs were more lucrative but then could not deliver to the end consumer due to transmission congestions.
Various solutions have been proposed in the prior art, some of which are explained below. One of the prior art discloses a power market management where a plurality of process streams having inter related process streams are established. However, these methods are intended to be used by an electric power dispatch centre or an exchange where price, day- ahead schedule, etc is optimized. This patent pertains to solving the problem of an independent system operator of managing the power trade/managing the share of different utilities. Here the key problem is to know the status of each generator, know the share of each utility in every generator, the position of each utility based on the availability, managing the transmission corridor and managing the power trade for utilities as independent exchange. Thus it provides solution to individual dispatch entity.
Another prior art discloses a decision support system for power trading. The decision support system here is intended to be used by generators to determine generation of electricity by integrating a wide variety of disparate data into a comprehensive information set that can be used to determine when, where, and how to generate electrical power. As explained here this invention is solving the problem from the generation side with different challenges.
One of the prior art discloses a demand forecasting system for the day-ahead using historical demand data, weather forecast for the next day and calculating the impact of day-of-the-week. This is using a time-series and regression based approach. This invention uses a similarity based approach for calculating weather impact by finding a day with similar weather in the past.
Another method provides a risk management system for electric utilities. A computer implemented risk-management system schedules the generating units of an electric utility while taking into consideration power trading with other utilities and the stochastic load on the utility system. The system provides the user multiple load forecasts and allows the user to vary the fuel price between different scenarios and different periods of planning horizon. However this system focuses on optimizing the price of fuel from generation side only. Thus this solution has limited approach.
Several models and methods have been tried to address individual component and problems associated with them. But none has looked at it in an integrated manner where linkages are strong. None of the method provides a solution to minimize the gap between demand and supply at distributor level. So in view of available prior art, there is need of a system which can optimize the generation, scheduling, distribution and supply among the entire network. The method is required which looks at the complete energy market in an integrated fashion and hence being able to optimize the system as a whole and to be able to increase available power at a reduced cost given the constraint. OBJECTIVES OF THE INVENTION
It is the primary object of the invention to provide a system and method facilitating forecasting, optimization and visualization of one or more parameters of energy data.
It is another object of the invention to provide a forecasting system facilitating multiple type of forecasting of energy data for one or more sectors of an industry.
It is another object of the invention to provide an energy optimization system facilitating multiple type of optimization of energy data for one or more sectors of industry.
It is yet another object of the invention to provide a visualization module in order to generate a visualization of gap thus observed in end results for better decision making.
It is yet another object of the invention to provide a risk measuring module configured to measure a risk factor associated with the energy data.
SUMMARY OF THE INVENTION
The present invention discloses an intelligent system to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data. The system comprises of a data gathering means configured to gather the energy data from one or more components distributed throughout the network, a data sorting module configured to categorize said energy data and transmit it to one or more sub-systems for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement and a network of subsystems, to receive categorized data for generating plurality of end results. The network further comprises of a forecasting sub system configured to process the categorized data and related constraints for the purpose of one or more type of forecasting for a particular parameter by applying one or more type of forecasting operations and an energy optimization sub system configured to process the categorized data and related constraints by means of its data processing components and generate one or more optimal solution for one or more parameters by using a dynamic optimization technique. The system further comprises of a communication channel configured to integrate the sub systems distributed throughout the network and enable the communication between them, the communication channel further enables an on-demand exchange of energy data and the end results amongst these sub systems and an output generation unit configured to communicate with one or more sub systems and captures the end results for further post-processing. The system further comprises of a data visualization module configured to further post-process the end result of one or more sub systems by using a suitable visualization methodology in accordance with the particular end result, such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry.
The present invention also discloses an intelligent method to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data. The method comprises of steps of gathering energy data from one or more components distributed throughout the network, categorizing said energy data and transmit it to one or more subsystems for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement and generating plurality of end results by receiving categorized data. The generation further comprises of processing the categorized data and related constraints for the purpose of one or more type of forecasting for a particular parameter by applying one or more type of forecasting operations and for energy optimization by means of one or more data processing components and generating one or more optimal solution for one or more parameters by using a dynamic optimization technique and processing the categorized data and related constraints. The method further comprises of integrating the sub systems distributed throughout the network and enabling the communication between them, the integration further enables an on-demand exchange of energy data and the end results amongst these sub systems, capturing the end results of one or more sub systems for further post processing and post-processing the end results of sub systems by using a suitable visualization methodology in accordance with the particular end result, such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry. The present invention also discloses a forecasting system to determine one or more types of forecasting results associated with energy for one or more sector distributing said energy. The system comprises of an input capturing device configured to capture data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module, a data filtration module configured to pre-process the data by using one or more data filtration techniques to remove an unwanted data and a processor configured to determine by way of its embedded algorithms one or more types of forecast results for a pre-defined time period. The processor further comprises of one or more sub- processing modules configured to apply one or more forecasting operations, such that the operations are applied with respect to the data thus captured and a rule based engine configured to perform one or more post processing steps over the forecast results by using predefined rules to further correct said forecasted results, the system further comprises of an output generation module configured to generate one or more types of corrected forecast results at one or more level associated with the energy distribution sector with respect to each source system.
The present invention also provides a forecasting method to determine one or more types of forecasting results associated with energy for one or more sector distributing said energy. The method comprises of steps of capturing data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module, pre-processing the data by using one or more data filtration techniques to remove an unwanted data and processing the data to determine by way of embedded algorithms one or more types of forecast results for a pre-defined time period. The processing further comprises of applying one or more forecasting operations, such that the operations are applied with respect to the data thus captured. The method further comprise of steps of post-processing the forecast results by using predefined rules to further correct said forecasted results and generating one or more types of corrected forecast results at one or more level associated with the energy distribution sector with respect to each source system.
The present invention also discloses a system to facilitate energy optimization with respect to one or more constraints for an industry type for a utility purpose. The system comprises of an integration module configured to integrate one or more source systems distributed across a network, a data capturing device configured to pull data from the source systems thus integrated throughout the network, the data capturing device further comprises of a constraint fetching module configured to take into account the constraints associated with the data thus pulled for a particular industry type and a processing engine configured to process the data in combination with the constraints to determine an optimized solution with respect to said constraint. The processing engine further comprises of an optimization module configured to form an objective function of said constraint by running a space reducing robust mechanism for reducing errors while determining the optimal solution. The system further comprises of an output generation module configured to generate the optimal solution with respect to the data thus captured, such that the optimal solution are used for one or more utility purpose.
The present invention also discloses a method to facilitate energy optimization with respect to one or more constraints for an industry type for a utility purpose. The method comprises of steps of integrating one or more source systems distributed across a network, pulling data from the source systems thus integrated throughout the network and taking into account the constraints associated with the data thus pulled for a particular industry type. The method further comprises of steps of processing the data in combination with the constraints to determine an optimized solution with respect to said constraint. The processing further comprises of forming an objective function of said constraint by running a space robust mechanism for reducing errors while determining the optimal solution and generating an optimal solution with respect to the data thus captured, such that the optimal solution is used for one or more utility purpose.
BRIEF DESCRIPTION OF DRAWINGS
Figure 1 illustrates the architecture of intelligent system in accordance with an embodiment of the invention.
Figure 2 illustrates a flow chart towards forecasting, optimization and visualization energy data in accordance with an embodiment of the invention.
Figure 3 illustrates the architecture of forecasting system in accordance with an alternate embodiment of the invention. Figure 4 illustrates the architecture of energy optimization system in accordance with an alternate embodiment of the invention.
Figure 5 illustrates different type of forecasting and optimization by exchanging data in accordance with an exemplary embodiment of the invention.
Figure 6 illustrates features of optimization in accordance with an exemplary embodiment of the invention.
Figure 7 illustrates the flow of data in a network in accordance with an exemplary embodiment of the invention.
DETAILED DESCRIPTION OF THE INVENTION
Some embodiments of this invention, illustrating its features, will now be discussed:
The words "comprising", "having", "containing", and "including", and other forms thereof, are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the listed item or items.
It must also be noted that as used herein and in the appended claims, the singular forms "a", "an", and "the" include plural references unless the context clearly dictates otherwise. Although any systems, methods, apparatuses, and devices similar or equivalent to those described herein can be used in the practice or testing of embodiments of the present invention, the preferred, systems and parts are now described. In the following description for the purpose of explanation and understanding reference has been made to numerous embodiments for which the intent is not to limit the scope of the invention.
One or more components of the invention are described as module for the understanding of the specification. For example, a module may include self-contained component in a hardware circuit comprising of logical gate, semiconductor device, integrated circuits or any other d iscrete component. The module may also be a part of any software program executed by any hardware entity for example processor. The implementation of module as a software program may include a set of logical instructions to be executed by the processor or any other hardware entity. Further a module may be incorporated with the set of instructions or a program by means of an interface.
The disclosed embodiments are merely exemplary of the invention, which may be embodied in various forms.
The present invention provides a system and method for processing energy data and generating one or more end results in a network. These results may relate to forecasting, optimization and visualization of one or more parameters associated with data. While processing the data the system also takes into one or more constraints and generates results accordingly. The data is gathered from plurality of sources or components distributed throughout the network. The data is pre-processed and the end results are post processed in order to further provide visualization of the data which may provide suggestive measures be used by one or more industry.
In accordance with an embodiment, referring to figure 1, the present invention comprises of an intelligent system (100) to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network. End results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data. The intelligent system (100) further comprises of a data gathering means (102) configured to gather the energy data from one or more components distributed throughout the network and a data sorting module (104) configured to categorize said energy data and transmit it to one or more sub-system (106) for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement.
The intelligent system (100) further comprises of a network of subsystems (106) to receive categorized data for generating plurality of end results. The network of subsystems (106) further comprises of an energy optimization sub system (108) and a forecasting sub system (110) configured to process the categorized data and related constraints.
The intelligent system (100) further comprises of a communication channel (112) configured to integrate the sub systems distributed throughout the network and enables the communication between them, an output generation module (1 14) a data visualization module (1 16).
In an embodiment of the invention, the integrated system (100) providing forecasting, optimization and visualization of one or more parameters associated with energy data may include but is not limited to electrical energy, water supply.
In an embodiment of the invention, the end results of forecasting, optimization and visualization depicts suggestive measures on considering one or more constraints associated with said data, said constraints may include but is not limited to power availability, unit price of power in exchange and grid, price and volume constraint for each source, grid stability related constraints, technical constraints relate to running the power plants, reliability requirement of each node in the network, to power surplus, power short fall estimation, contract.
The data gathering means (102) gathers the energy data from one or more components distributed throughout the network (step 202 of figure 2). These components may also be referred to as source components or source systems. The energy (electric energy) data which is thus gathered is common and this data should be categorized so that it may be used and fed for the purpose of forecasting, optimization and visualization. The data sorting module ( 104) performs this task of categorizing this data and then transmitting it to the respective subcomponent (steo 204 of figure 2). This data gathering module (102) acts as a source system for all the sub components processing the data.
The present invention will now be explained by way of various sub-systems and processing of the energy data through them for generating end results (step 206 of figure 2).
Referring to figure 3, the forecasting system (300) (also referred as forecasting sub system) further comprises of an input capturing device (302) configured to capture data from one or more source systems, a data filtration module (304) and a processor (306).the processor further comprises of one or more sub-processing modules (308) configured to apply one or more forecasting operations; The forecasting system (300) further comprises of a rule based engine (310) configured to perform one or more post processing steps over the forecast results, an output generation module (312) configured to generate one or more types of corrected forecast results.
The forecasting system (300) implements a method to determine one or more types of forecasting results associated with energy for one or more sectors distributing said energy (step 208 of figure 2). The energy further includes electrical energy. One or more types of forecast results further comprises of results for Electricity demand , forecasting, UI (Unscheduled Interchange) Forecasting and IEX (Exchange Price) Forecasting.
The forecasting system (300) further comprises of an input capturing device (302) configured to capture data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module. The data captured by an input capturing device (302) further comprises of data related to historical electricity/energy consumption and distribution data, weather data, grid-wise data, population and past demand in order to determine electricity demand forecasting, UI(unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
By way of specific example, the data captured by an input capturing device (302) in order to determine electricity demand forecasting may include but is not limited to Historical Electricity Demand, Historical Weather, Forecasted Weather, Weather Variation Range, Holiday List, Population Growth Rate, Energy Consumption Growth Rate, GDP Growth Rate, Industrial Output Growth, Agricultural Growth Rates, etc.
Historical load data includes details on power cuts to calculate Unrestricted Peak Demand (UPD). If this data is available at a segment or feeder level, a breakdown of demand at these levels can be obtained. Weather Data includes Temperature, Humidity, Rainfall, Wind Speed and Cloud cover. Cluster/ location details at a desirable level, Calendar Data including most of the calendar effect details which are commonly available, if there are area specific events/calendar effects that also included, Events details and list and details of power sources (PPAs, Own generation details etc.).
By way of specific example, the data captured by an input capturing device (302) in order to determine UI (Unscheduled Interchange) forecasting includes Grid-wise Historical drawl data aggregated over all distributors in the grid, Grid-wise schedule aggregated over all distributors, Grid wise generation, Historical UI values for the grid, Historical weather for representative locations in the grid, weather forecast for representative locations in the grid, Holiday list.
By way of specific example, the data captured by an input capturing device (302) in order to determine IEX (Exchange price) forecasting includes Historical Exchange Price, Historical weather for representative location, condition of generating units that feeds in power to this region, the demand pattern (fluctuations) for every beneficiary in this region, weather forecast for representative locations in the grid, Holiday list.
The data filtration module (304) of forecasting system (300) is configured to pre-process the data by using one or more data filtration techniques to remove an unwanted data. There can be errors in the captured data due to variety of reasons such as bad data capture practices, distribution or transmission failure etc. These aberrations in the data may result in an incorrect relation between input and output. So in order to improve the accuracy of the model and a better understanding of the relationship, pre-processing step is carried out to clean the captured data.
Data preprocessing step is carried out by using one or more data filtration techniques. The data filtration techniques used may include use of an adaptive median filter to filter the data. An adaptive median filter applied for each time block to mark unusually high/low Demand/Weather/UI/Frequency values. If the data is suspected, it is either discarded or a suitable replacement is found based on historical data/basic forecast. Finally the filtered data is used further in forecasting.
The processor (306) is configured to determine by way of its embedded algorithms one or more types of forecast results for a pre-defined time period. The forecast results for predefined time period ranges from 15 minutes to 15 years. The processor further comprises of one or more sub-processing modules (108) configured to apply one or more forecasting operations, such that the operations are applied with respect to the data thus captured. The forecasting operations further comprises of a regression based approach, additive splines, parameter shrinkage, grid-level demand/schedule/generation and generic algorithms or a combination thereof in order to determine electricity demand forecasting, UI(unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
In accordance with an exemplary embodiment, the forecasting operation steps referring to Electricity demand forecasting includes but not limited to following steps. Regression base approach is used to relate the future demand to- historical demand, weather, events, planned schedules etc. The regression tries to identify relationships between future demand and these parameters. These relationships are then used to project future demand. Additive Splines are used to capture non linearity with weather and Demand. Parameter shrinkage is done to avoid over-fitting. Separate model is built for each time block. Adaptive mean error adjustment is done to capture local effects. Holidays captured through a factor based model Probabilistic density forecasting is used to capture forecasting uncertainty. In some cases, an ensemble of multiple methods was used to get a better result. Regression based approaches and neural network is used with different inputs to explore different kinds of relationship between inputs and the demand. Further an ensemble method is used to combine these relationships into one that gives a better prediction.
Further in accordance with an exemplary embodiment, the forecasting operation steps referring to UI (unscheduled Interchange) forecasting includes but not limited to Regression based approach and use of additive splines.
In accordance with an exemplary embodiment, the forecasting operation steps referring to IEX (Exchange price) forecasting includes but not limited to: Applying Regression based approach using historical IEX data. Current generation Injection schedule is captured in the model. Current grid shortage is captured in the model. Buy and sell bidding is captured by using "sigma" factors (Square root of average deviationA2 between forecast and actual). This is further used to optimize the range that determines the sale and buy bids. Bid clearance and buy-sell price gap is simultaneously optimized. Alternatively, in some cases, genetic algorithms are also used for the same.
The forecasting operations are performed by the sub processing modules such that the operations are applied with respect to the data thus captured and the required forecasting result type such as electricity demand forecasting, UI (unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
The rule based engine (310) is configured to perform one or more post processing steps over the forecast results by using predefined rules to further correct said forecasted results. The rule based engine is configured to perform one or more post processing steps further comprises of a use of Kalman filter in order to sort the errors in historical data.
By way of specific example, the data post processing steps includes making use of specific rules developed for each utility for example, Correction for Agricultural Load Change, Correction for planned events etc. Further Post-processing techniques also includes use of Kalman Filter in order to do an error correction and thus eliminate un-wanted noise in forecasting. This works as a local model to correct local deviations between forecasted and actual parameters. The rules are used to override or correct the system forecasts. There are two kinds of corrections. In the first way, it is done by the system itself where it looks at the historical errors and corrects itself to get closer to the actual. In the second way, it is done manually, where the user can change the forecast based on their understanding of the market. The system takes the inputs from the users in terms of whether the demand will go up or down in a block or set of contiguous blocks and ensures that the demand pattern is maintained while increasing or reducing the demand.
The output generation module (312) of forecasting system (300) is configured to generate one or more types of corrected forecast results at one or more levels associated with the energy distribution sector with respect to each source system. The one or more types of corrected forecast results further comprises of a point forecast, a range or a probability distribution or a combination thereof with respect to the electricity demand forecast, UI/Frequency Forecast at 15 minute interval with respect to IEX (Exchange Price) Forecasting and Lower and upper limit for exchange Price at 15 minute interval for the specified region with respect to IEX (Exchange Price) Forecasting.
One or more levels associated with the energy distribution sector further comprises of an aggregate distributor level, station level, sub-station level, feeder level or a combination thereof. In accordance with an exemplary embodiment, the corrected forecast results generated with respect to electricity demand forecasting includes electricity Demand Forecast for the distributor at multiple horizons. More particularly it is for an hour ahead, day-ahead, day- week ahead, months ahead, year ahead, and 3-20 years ahead. The output can be in the form of a point forecast, a range or a probability distribution. The forecasts can be generated at an aggregate distributor level, station level, sub-station level, feeder level etc. This forecast is used in Power Procurement to meet demand as well as having some flexibility to do power trading within the limits specified by the regulator.
By way of specific example, electricity demand forecast is generated as short term forecast and medium term and long term demand forecast. The short-term demand forecast module generates forecast for individual blocks (96 blocks of 15 minutes interval) at the following level of granularity including, day ahead, week ahead, max, min and volatility. It can be extended to generate this forecast at various level of aggregation across dimension based on geography and usage. For example, the user might want to generate area or cluster wise load forecast for industrial, domestic, commercials, government offices, Railways etc. This depends on the availability of data at the desired level.
In accordance with another exemplary embodiment, medium term demand forecast can be generated at hourly or peak and off-peak on a daily level. Daily Forecast of Peak Demand, Base Demand for coming 1 year is predicted. The forecast is divided into multiple blocks of a day such as Early morning demand, Late morning demand, Afternoon demand and Late evening demand. Alternatively, rather than a demand forecast for these blocks of a day, a range of demand for these blocks at a pre-specified confidence level is predicted which includes Most likely demand, As-high-as demand and As-low-as demand.
By way of an example, medium term demand forecast includes generating year ahead demand forecast for 15 minutes interval or 3 to 5 years demand forecast for 1 hour interval.
By way of an example, long term demand forecast includes generating demand forecast for 25 years at monthly interval.
In accordance with an exemplary embodiment, the corrected forecast results generated with respect to UI (unscheduled Interchange) forecasting includes Unscheduled Interchange or Grid Frequency Forecast at 15 minute interval to several days ahead. The forecast result can be used by the distributor to anticipate likely values of the grid frequency during the day leading to optimal planning, procurement, shutting down/ramping up of captive generation, etc.
In accordance with an exemplary embodiment, the corrected forecast results generated with respect to IEX (Exchange Price) forecasting includes lower and upper limit for exchange Price at 15 minute interval for the specified region. Further in exchange price forecasting, forecasting horizon can be from 15 minutes to several days ahead. The exchange price forecasting module predicts the market clearing price for multiple regions. The range in which the market clearing price will fall is predicted. This helps the user in deciding the bidding price for every block in a day. This forecast can be used by the distributor to anticipate likely values of the exchange price during the day leading to optimal planning, procurement, shutting down/ramping up of captive generation, etc.
In accordance with an exemplary embodiment, the demand forecast result decomposition is carried out by area, region or cluster. The forecast result can be further decomposed by type of customers wherein type of customers can be further categorized as migrated customers, own customers and open access customers. Further, custom grouping can be specified for example by per capita, load factor, density and diversity factor.
In accordance with an exemplary embod iment, features of the forecast result more particularly electricity demand forecast comprises of capturing non-linear relationship with weather parameters, Modelling load variation during a day, Modelling weekday, weekend variation, Capturing Holiday effect, applying adaptive correction to learn from past forecasting errors, providing facility to incorporate domain expertise input into forecast output, carrying out risk or variation quantification of the forecast errors in different situations, carrying out optimistic and Pessimistic Scenario Analysis for multiple weather conditions, taking into account Economic, Demographic or Geographic parameters like GDP Growth, Energy consumption per capita growth, population under electricity growth, industrial growth for long term forecast. In accordance with an exemplary embodiment, features of the forecast result more particularly UI (Unscheduled Interchange) Forecast comprises of capturing day of the week effect, capturing time of the day effect, capturing holiday effect, capturing non-linearity of weather parameters and analyzing risk with respect to any production unit failing or abnormal increase in demand.
In accordance with an exemplary embodiment, features of the forecast result more particularly IEX (Exchange Price) Forecast comprises of capturing day of the week effect, capturing time of the day effect, captures holiday effect, capturing non-linearity of weather parameters and analyzing risk of price and volume.
In accordance with an exemplary embodiment, features of the forecasting system, includes adaptable to new data and parameters, flexible to accommodate features including event database to capture special conditions, scale up and scale down forecast at overall and individual block levels, forecasting at different granularity, region or cluster level forecast, customer segment level forecast and simple to use and interactive modelling platform.
In accordance with an embodiment, the forecasting system (300) enables the user to use his domain knowledge or any region or date specific additional information to change the system generated forecast. The revised forecast is also stored in the system and used for any future analysis which becomes very useful in the case of special events that are not declared in advance. By way of an example, a strike declared by a particular political party or a local body election.
The system (100) now further comprises of a sub system to facilitate energy optimization (also referred to as energy optimization or simply optimization system) which is present in the network of sub-systems and is integrated with all the other sub-systems (step 210 of figure 2).
In accordance with an embodiment, referring to figure 4, the energy optimization system (400) for energy optimization with respect to one or more constraints comprises of an integration module (402) configured to integrate one or more source systems distributed across a network. The energy optimization system (400) further comprises a data capturing device (404) to pull data from the source systems throughout the network. The data capturing device (404) further comprises a constraint fetching module (506) configured to consider the constraints associated with the data pulled from the source systems.
The optimization system (400) further comprises of a processing engine (408) configured to process the data in combination with the constraints to determine an optimization solution with respect to said constraint. The processing engine (408) further comprises an optimization module (410) configured to form an objective function of said constraint by running a space robust mechanism for reducing errors while determining the optimal solution. The energy optimization system (400) comprises of an output generation module (412) configured to generate the optimal solution with respect to the data thus captured.
In accordance with an embodiment, the system (400) to facilitate energy optimization for an industry type, the industry type may include but is not limited to electrical energy, water supply.
The energy optimization system (400) comprises of said integration module (402) configured to integrate said one or more source systems distributed across network. The source system comprises of one or more forecasting systems, PPA, Hydel generation, Thermal generation, Gas generation, Bilateral, Exchange, UI.
In an embodiment of the invention, the data capturing module (404) is configured to pull the data from the source systems thus integrated throughout the network. The data pulled from the source systems may include but not limited to the data related to price and availability of each of power sources, price of deviating from the agreed volume, reliability of each power source, corridor limit and transmission constraints, demand forecast, UI forecast, exchange price forecast, technical specifications on each type of power source.
In an exemplary embodiment of the invention, the data capturing module (404) pulls the data from the source systems in real-time.
In accordance with an embodiment, the data capturing device (404) further comprises of the constraint fetching module (406) configured to take into account the constraints associated with the data pulled for particular industry type. The constraints associated with the data may include but not limited to power availability, unit price of power in exchange and grid, price and volume constraint for each source, grid stability related constraints, technical constraints relate to running the power plants, reliability requirement of each node in the network. Further, the constraints associated with the data may include but not limited to power surplus, power short fall estimation, contract.
In yet another embodiment of the invention, the energy optimization system (400) comprises of the processing engine (408). The processing engine (408) is configured to process the data in combination with the constraints to determine optimized solution (not shown) with respect to the constraint. The processing engine (408) further comprises the optimization module (410) configured to form objective function of the constraint by running said space reducing robust mechanism (which is a dynamic optimization technique) for reducing errors while determining the optimal solution. The constraint reducing robust mechanism (not shown) further comprises of a mathematical programming approach, a heuristic approach or a combination thereof.
The output generation module (412) generates the optimal solution with respect to the data thus captured, such that the optimal solution is used for one or more utility purpose.
The optimal solution further comprises of a short term energy portfolio optimization and a medium term energy portfolio optimization.
The energy optimization system provides following end results:
► Optimal evaluation of quantum of contract for sale or purchase
► Provides an optimized solution based on the following constraints
0 Contracts, their obligations and rules
° Power Surplus / Short fall estimation
° Spot and Day ahead market availability and rate
Different types of energy optimization are discussed below (depending upon the input data):
Short term Portfolio Optimization: The energy optimization system provides the solution to the short term portfolio optimization in a unified unit commitment problem as well as financial decisions relating to short term bilateral, PPP, Day ahead forward and spot market. Price as well as volumetric risk is considered for risk management. The system integrates an Industrial book (details of the generation units owned and managed by the utility) along with a Trade book (details of all external sources of power with details of terms and price associated with each agreement) for the optimal decision making process.
For example, once the utility has an estimate of the demand, it can fulfil that demand by a combination of supply from different sources (internal as well as external detailed in the Industry and Trade Book). Each of these sources are associated with an entitlement of power for every time- block and cost for deviating from that entitlement. There are several other constraints that need to be taken care of while deciding on an optimal mix of power from different sources.
The technical constraint of the power plant, minimum/maximum down-/up-time, start-up costs, operational cost as well as uncertainty of availability is taken into consideration. For the optimal bidding mechanism, the behavior of the competitors is also taken into account in the game theoretic framework. Optimal strategy is developed for different hour of the day such as peak and off-peak hours. Optimal strategy also takes into account the changing load shape depending on the season.
The salient features of the short-term optimization module are as following:
(a) It pulls in the demand forecasting (step 502 in figure 5) for the next day from a demand forecasting module (forecasting system). Several forecasts may be generated for a day by providing different inputs or by revising the forecast generated by the system (step 504 of figure 5).
(b) Takes all possible avenues of power procurement/ selling to come up with an optimal decision (e.g. how to utilize existing PPAs, how much to buy/ sell at exchange, generation scheduling at captive power plants, bilaterals and banking).
In accordance with an embodiment, the energy optimization further provides a flexibility to the user to change any of these data before running the dynamic optimization technique. For example, if a particular plant is shut down for maintenance, the user can remove it from the optimization. Similar changes can be done for trade book, if required.
The system automatically connects to SLDCs and fetches the real-time running capacity of each unit of the plants in the industry book. Based on this real-time data, it predicts the running capacity of the units for the next day. On top of this, the user can change the capacity at a block level based on any additional information.
It also connects to the forecasting system and fetches exchange and UI forecasting from that system as these are essential components used in portfolio optimization.The exchange forecasting can also provide a range rather than a point forecast. This will help the utiltiy in deciding on the bidding strategy.
It further provides an interface to capture the transmission constraint at the LDC level and exchange level. The key objective is to provide a decision that minimizes the overall cost of power procurement. The decisions include how much power should be taken from each source, how to do a trade-off between exchange and UI, how to schedule the generation plants etc. The energy optimization system also provides the total cost for this portfolio and the risk measures around this decision.
The inputs to the optimization system include several forecasts (demand forecast, price forecast, availability forecast etc.) and hence the quality of the solution is dependent on the quality of these inputs. However, it is obvious that these inputs will have uncertainty associated around it. If the optimization is done using these estimates as crisp inputs without understanding the uncertainty, the optimization solution may not remain optimal (or even feasible) if these inputs change. Therefore, there is a need of a solution that is more robust to these changes in the inputs. The energy optimization system uses state-of-the-art Robust Optimization approach to take care of variability in the inputs and uncertainty in the environment to provide a more robust decision. The system uses the following approach to introduce robustness in the system. The system has a sensitivity analysis to enable users to do what-if analysis.
1. The system generates multiple estimates to capture different scenarios. Based on these multiple estimates, a most-likely and two extreme estimates are generated. 2. The optimization problem is converted into a multiple objective optimization problem. The first objective is to optimize the most-likely case. Whereas the other two objectives intend to minimize the loss in worst case and maximize the gain in the best case situation.
The energy optimization system (400) further integrates with external systems for workflow to enable communication of decisions between teams (back-office, middle-office and trading teams) and getting feedback. The energy optimization system is automated enough to minimize human error, at the same time, can incorporate feedback provided by experts. This system takes care of technical constraints (swing available on a PPA, UI limit or idling and switching constraints on power plants etc.) as well as business constraints (high reliability for special areas, distribution of power between urban and rural areas etc.).
Medium Term Portfolio Managpnent
In this type of optimization, the system helps the user to make decisions on buying or selling of power in the medium term (upto 1 year ahead). It provides recommendation on buying/ selling RTC power and block-wise power. Furthermore, since we are considering a longer time horizon the system does not provide one recommendation for each period. It provides multiple recommendations based on the confidence level.
The key features of the medium and long term optimization modules are as follows: a. Takes the inputs from the forecasting system of the intelligent system. b. Takes inputs from trade book and Industry book regarding existing agreements and available power from other sources. c. Calculate the Position Gap at different levels (RTC, Block-wise) d. Find the optimal mix of trade opportunities at different sources
In an exemplary embodiment of the invention, the optimum solution is implemented by two methods- mathematical programming approach by way of combining integer linear programming problem and heuristic method in order to solve the problem in multiple steps. The mathematical programming approach is explained in way of an example mentioned below. All these steps are performed by the processor thus present in the energy optimization system:
The integration module configured to integrate said one or more source systems comprising PPA, Hydel generation, Thermal generation, Gas generation, Bilateral, Exchange, UI.
Exchange <0□ sell to exchange
UI<0□ underdraw from the grid
The constraint fetching module considering the constraints associated with the data pulled, the constraints associated with the data may be presented as follows:
0. The demand needs to be met. For each time-block... Sum(power procured from all sources) = Demand - power_cut. . This is adjusted based on the transmission and distribution losses.
1. Power cut distribution. The power cut at each time-block needs to be distributed across the distribution network to ensure reliability of the network. This is done by a combination of two constraints. For all areas with 100% reliability requiremet power_cut = 0. For the rest the power cut is propotional to (1 -reliability index).
2. For each block, MAX DRAWL ALLOWED FROM GRID (ppa, exchange, ui, bilateral) <= Max allowed for grid (corridor constraint). This will take care of any congestion in the corridor.
3. For all blocks put together, Ensure drawl from ui is not more than x% of drawl from PPA, EXCHANGE & BILATERAL. This is a regulatory constraint and the maximum amount that one can overdraw from the grid at different frequency.
4. ppa_drawl <= ppa drawl Limit (entitlement)(zero deviation allowed above it, but can plan in advance to draw less).
5. ppa_drawl_limit = under_drawl_from_ppa + ppa drawl (variable from 4th constraint) 6. Drawl from PPA is >= deviation*ppa drawl Limit(entitlement)*reliability of_unit. This deviation is different for different agreements.
7. Exchange is limited on either side by the respective MAX CAPACITIES. These MAX_CAPACITY can be estimated using the historical volume transaction and the corridor limit of a utility. 8. Availability of power plants (non-hydel). If the unit is not down, the maximum it can generate is its max capacity. (non_hydel) Powerplant drawl <= MAX CAPACITY
9. Partial generation of the power plant. The output from a plant can be reduced to x% of the max capacity. (non_hydel) Powerplant drawl >- x%*MAX CAPACITY. There are two ways of ensuring this constraint. First, when the power plant can be run at any stage between x% and 100%. Second, where the power plant can be run only on specific stages (100%, 80%, 60%, 0% etc.). In the second case an indicator variable is used to indicate the stage at which the power plant is running and the abrupt change constraint is introduced on any change in the stage.
10. Abrupt change constraint: Powerplant generation cannot be changed very often, especially for thermal units. Once a change in generation is made, it needs to be the same for atleast 4 hours. Once it is shut down, it needs to be down for a day. For a gas or liquid plant, these changes can be made more frequently.
1 1. Full drawl from Hydel plants (limited by its RUNNING CAPACITY)
12. Availability of power plant (hydel): there can be two kinds of hydel plants (reservoir based and run of the river based). In a reservoir based plant, the output can be controlled to a maximum of level of water, max capacity of the unit. In case of run-of-the-river unit, the running capacity will be in the same vicinity for consecutive days.
13. All variables associated with source type, except exchange and ui are positive.
The processing engine processes the data in combination with the constraints to determine optimized solution with respect to the constraint. The processing engine further comprising the optimization module forms objective function of the constraint by running a space reducing robust mechanism (not shown) for reducing errors while determining the optimal solution. The objective function for exemplary purpose considered are- Minimize (variable_cost_ppa * ppa_drawl + fixed_cost_ppa*under_drawl_from_ppa + cost_hydel*hydel_drawl + cost_non_hydel * non_hydel_drawl + cost_exchange * exchange drawl + cost ui * ui_drawl ) - penalty * power_cut
In addition to the constraints presented in detailing the exemplary embodiment, other rapid change constraint are considered which may include but not limited to plants to run at full capacity mandatorily, plant operational hours.
The optimum solution is presented by implementing Heuristic Approach to the problem. The heuristic approach is performed in sequence.
In heuristic approach for each block:
1) The data is organized as a table with source name, source type, minimal enforced drawl as per the agreement, price, overall power available from it, power available for optimization and recommended volume which needs to be filled in.
2) First the enforced drawl is allocated as per the agreement with each source.
3) Power left after allocating the enforced drawl is filled in as "power available for optimization".
4) The table is sorted by price in ascending order.
5) "power available for optimization" is allocated starting from the lowest price as per the sorted order, till the demand for the block is either met or the supply is exhausted.
6) For sources with abrupt change constraint, step 5 is done so as to ensure that the constraint doesn't get violated. In this case, steps 3,4,5 are repeated iteratively to ensure that maximum power gets, utilized from sources with lower prices and at the same time ensuring that the abrupt change constraint doesn't get violated. The iterative process can be stopped when the final cost between two iterations is lower than a pre-specified threshold.
7) In the above process, if there is extra available power after the demand is met, the power needs to sold to either exchange or ui, depending on their predicted costs. If cost of supply is lower than the cost of sale (either ui or exchange), then appropriate amount is allocated for drawl from such a source in the recommended volume and is sold to either ui or exchange first (by storing it as a negative allocation), depending on which is more expensive, so as to maximize the profit margin.
8) All the other constraints mentioned in the mathematical formulation are ensured by coding them as conditional statements where they are appropriate.
The optimal solutions are used for utility purpose selected from a group of Prioritize across contracts, spot market and captive generation, Amount of volume to be exercised across contract, how much to buy or sell (depending on the supply Available) from day-ahead exchange and spot market, how much to generate if at all at any given point in time, given current status of various generation units and their marginal cost and the spot price, Prioritize Ramp-up & Ramp Down, which unit to shut down in case of excess capacity or a combination thereof.
Depending on the type of problem such as day-ahead, medium term, short term etc, the time unit is changed from 15 minute block to off-peak, on-peak to month. The output is presented as volume of power to be procured at each time unit, buy or sale of power and bidprice for exchange, how to run internal generation units, how much power to surrender, how much buffer to keep, what is the risk associated with the solution.
Referring to figure 6, the energy optimization system shares data from almost all the sub systems and modules present in the network of sub systems in order to generate end results. The shared data further includes price forecast, available power, contracts, committed power, generation cost schedule, gap, node reliability availability, risk appetite, demand forecast etc.
The network of subsystem further comprises of modules like network planning module, position mapping module and power flow module.
The network planning module is configured to capture the availability of supply from different sources, Identifying the risk attached (reliability) to the expected supply of electricity from different sources and providing scenario based avai lability of electricity from different sources. The position mapping module is configured to aggregate the demand and supply in the market to understand the position gap at different granularity and evaluating whether it is for the entire day or for different blocks of the day, whether it is at an aggregate level or at a sub-station/ feeder level and provides the position gap.
The power flow subsystem is further configured to provide optimal schedules for generation units over a 24 hour period which in turn minimizes operating cost of generation plants. By way of an example, constraints included here are load balance, spinning reserve, down time limits, C02 emissions and ramp rate limits.
The data (or end results) from these three or more modules (network planning, power flow and position mapping) is again shared by energy optimization system for further processing and processing end results for the energy optimization system (as shown in step 508 of figure 5).
The intelligent system (100) now further comprises of the communication channel (1 12) configured to integrate the overall system with all the sub-systems (and modules) and external components (step 212 of figure 2). This integration is performed by communicating with individual integration module which is coupled with each sub systems (forecasting and optimization). Because of this integration, these sub systems may exchange data with each other in case of any demand. For example, the energy optimization system (or sub system) is pulling out data from the forecasting system (sub system).
Lastly, in order to collect the end results from the entire sub systems (forecasting and optimization), the system (100) comprises of the output generation unit (1 14) which is again transmitting the end results to further components for further post processing (step 214 of figure 2).
The data visualization module (1 16) further executes steps of post-processing the end results of sub systems (106) by using a suitable visualization methodology in accordance with the particular end result such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry (step 216 of figure 2). The suggestive measures may be taken (depending on the infrastructure and data availability), at utility level, sub-station level, feeder level or even at a customer level.
In an exemplary embodiment of the invention, the visualization further comprises of a color code determining a size of a position gap on a continuous color gap.
The data visualization module (1 16) further comprises of a report generation module configured to generate reports that enable the users to analyze accuracy of the forecast result at several granularities. The system provides a very transparent way to check the accuracy of the forecast result on a particular day, over a week or over last 100 days. The forecasting system also enables the user to separate errors due to weather forecasting error and demand forecasting errors. On a daily basis, the system compares weather forecast versus actual weather to enable this analysis. MAPE (mean Absolute percentage Error) on a daily basis
✓ MAPE Distribution Report : % of days with MAPE <=2%, % of days with MAPE between 2% and 4%, % of days with MAPE between 4% and 6%, % of days with MAPE between 6% and 8% and % of days with MAPE >8% y Block-wise actual MAPE report as a heat map f Model tracking Report y Weather forecasting accuracy report f Visual comparison of demand forecast and actual for a particular day. f Other Customized Report can be frozen during the requirement analysis phase.
The intelligent system (100) also allows the users to separate errors due to weather forecasting error and demand forecasting errors. On a daily basis, the system compares weather forecast vs actual weather to enable this analysis y Aggregated view of demand and supply to understand the gap f An enterprise wide snap shot of position gap at varying levels of granularity y PowerMatics powerful visualization capability helps drill down todesired levels
The intelligent system (100) comes bundled with state of the art powerful visualization capability which allows the user to analyze position gap visually where the color code determines the size of the position gap on a continuous color scheme. The system (100) facilitates multidimensional visual charts which are easily comprehensible. The charts can be displayed on computer screens, large scale displays, iPads etc.
In an exemplary embodiment of the invention, the intelligent method further comprises step of measuring risk to further determine one or more factor leading to a risk with respect to energy data.
One view of utility trades and risk Composed of
o Trade book
o Industrial book
o Electricity Transaction book
These books allow integrating all portfolios to ensure you have comprehensive view of your risk.
These are further divided it into the three broad categories.
X Trading book: comprising of all power contracts both
OTC and Exchange Traded
X Industry Book: industrial book can be divided into two
part
o Portfolio of generation Units,
o Distribution and transmission infrastructure.
Trade Book comprises of:
► Captures utility's "Buy contracts"
► Long term contracts and their associated conditions
► Short term contracts along with conditions, trading margin, open access and miscellaneous charges > Exchange contracts, conditions, energy and transmission charges
Industry Book comprises of:
> Captive generation plant's configuration
> Individual Unit's parameters and configuration
> Fuel Formula / Market pricing
Optimal fuel mix generation
As shown in figure 5, (step 506), the data from electricity transaction book, trade book and industry book after processing goes to the energy optimization system.
In accordance with an embodiment, referring to figure 7, the intelligent system (100) transmits end results of all the sub systems throughout the network by using various communication sources like intranet, internet etc. The transmitted data could be easily accessed or controlled over PC, tablets and request for such data could also be sent through SMS emails etc. Various purposes in which the end results could be further utilized are for contract and invoice, demand forecast and trading.

Claims

WE CLAIM:
1. An intelligent system to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data, the system comprising: a data gathering means configured to gather the energy data from one or more components distributed throughout the network;
a data sorting module configured to categorize said energy data and transmit it to one or more sub-systems for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement;
a network of subsystems, to receive categorized data for generating plurality of end results, the network further comprising:
a forecasting sub system configured to process the categorized data and related constraints for the purpose of one or more type of forecasting for a particular parameter by applying one or more type of forecasting operations; an energy optimization sub system configured to process the categorized data and related constraints by means of its data processing components and generate one or more optimal solution for one or more parameters by using a dynamic optimization technique;
a communication channel configured to integrate the sub systems distributed throughout the network and enable the communication between them, the communication channel further enables an on-demand exchange of energy data and the end results amongst these sub systems;
an output generation unit configured to communicate with one or more sub systems and captures the end results for further post-processing; and
a data visualization module configured to further post-process the end results of one or more sub systems by using a suitable visualization methodology in accordance with the particular end result; such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry.
2. The intelligent system of claim 1, wherein the intelligent system further comprises of a risk measuring sub system configured to determine one or more factor leading to a risk with respect to energy data.
3. The intelligent system of claim 1, wherein the network of sub systems further comprises of a network planning module, a position mapping module and a power flow module.
4. An intelligent method to provide forecasting, optimization and visualization of one or more parameters associated with energy data in a network, while end results of said forecasting, optimization and visualization further depicts suggestive measures on considering one or more constraints associated with said data, the method comprising steps of:
gathering energy data from one or more components distributed throughout the network;
categorizing said energy data and transmit it to one or more sub-systems for further processing, to initiate a related forecasting, optimization and further visualization as per an industry requirement;
generating plurality of end results by receiving categorized data, the generation further comprising:
processing the categorized data and related constraints for the purpose of one or more type of forecasting for a particular parameter by applying one or more type of forecasting operations;
processing the categorized data and related constraints for energy optimization by means of one or more data processing components and generating one or more optimal solution for one or more parameters by using a dynamic optimization technique; integrating the sub systems distributed throughout the network and enabling the communication between them, the integration further enables an on-demand exchange of energy data and the end results amongst these sub systems;
capturing the end results of one or more sub systems for further post processing; and post-processing the end results of sub systems by using a suitable visualization methodology in accordance with the particular end result;
such that the visualization of end results is further analyzed by analysis module in order to generate suggestive measures for any gap thus found for one or more sector of an industry.
5. The method of claim 4, wherein the visualization further comprises of a color code determining a size of a position gap on a continuous color gap.
6. The method of claim 4, wherein the intelligent method further comprises of a step of measuring risk to further determine one or more factor leading to a risk with respect to energy data.
7. A forecasting system to determine one or more types of forecasting results associated with energy for one or more sector distributing said energy, the system comprising: an input capturing device configured to capture data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module;
a data filtration module configured to pre-process the data by using one or more data filtration techniques to remove an unwanted data;
a processor configured to determine by way of its embedded algorithms one or more types of forecast results for a pre-defined time period, the processor further comprising:
one or more sub-processing modules configured to apply one or more forecasting operations, such that the operations are applied with respect to the data thus captured; a rule based engine configured to perform one or more post processing steps over the forecast results by using predefined rules to further correct said forecasted results; an output generation module configured to generate one or more types of corrected forecast results at one or more level associated with the energy distribution sector with respect to each source system.
8. The system of claim 1 and 7, wherein the energy further includes electrical energy.
9. The system of claim 1 and 7, wherein the data filtration is further performed by using an adaptive median filter to filter the data.
10. The system of claim 1 and 7, wherein the rule based engine configured to perform one or more post processing steps further comprises of a kalman filter in order to sort the errors in historical data.
11. A forecasting method to determine one or more types of forecasting results associated with energy for one or more sector distributing said energy, the method comprising: capturing data from one or more source systems distributed across a network, said source systems are integrated by means of one or more integration module;
pre-processing the data by using one or more data filtration techniques to remove an unwanted data; processing the data to determine by way of embedded algorithms one or more types of forecast results for a pre-defined time period, the processing further comprising: applying one or more forecasting operations, such that the operations are applied with respect to the data thus captured; post-processing the forecast results by using predefined rules to further correct said forecasted results; and
generating one or more types of corrected forecast results at one or more level associated with the energy distribution sector with respect to each source system.
12. The method of claim 4 and 11, wherein the energy further includes electrical energy.
13. The method of claim 4 and 1 1, wherein the data filtration techniques are further performed by using include an adaptive median filter to filter the data.
14. The method of claim 4 and 1 1, wherein one or more type of forecast results further comprises of results for Electricity demand forecasting, UI (Unscheduled Interchange) Forecasting and IEX (Exchange Price) Forecasting.
15. The method of claim 4 and 1 1, wherein the data captured by an input capturing module further comprises of data related to historical electricity/energy consumption and distribution data, weather data, grid-wise data, population and past demand in order to determine electricity demand forecasting, UI(unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
16. The method of claim 4 and 1 1, wherein the forecast results for pre-defined time period ranges from 15 minutes to 15 years.
17. The method of claim 4 and 1 1, wherein the forecasting operations further comprises of a regression based approach, additive splines, parameter shrinkage, grid-level demand/schedule/generation and generic algorithms or a combination thereof in order to determine electricity demand forecasting, UI(unscheduled Interchange) forecasting and IEX (Exchange price) forecasting.
18. The method of claim 4 and 1 1, wherein one or more level associated with the energy distribution sector further comprises of an aggregate distributor level, station level, sub-station level, feeder level or a combination thereof.
19. The method of claim 4 and 1 1, wherein the one or more types of corrected forecast results further comprises of a point forecast, a range or a probability distribution or a combination thereof with respect to the electricity demand forecast, UI/Frequency Forecast at 15 minute interval with respect to IEX (Exchange Price) Forecasting and Lower and upper limit for exchange Price at 15 minute interval for the specified region with respect to IEX (Exchange Price) Forecasting.
20. The method of claim 4 and 1 1 , wherein post processing steps of the forecast results is performed by using a kalman filter in order to sort the errors in the historical data.
21. An integrated system to facilitate energy optimization with respect to one or more constraints for an industry type for a utility purpose, the system comprising:
an integration module configured to integrate one or more source systems distributed across a network; a data capturing device configured to pull data from the source systems thus integrated throughout the network, the data capturing device further comprises of a constraint fetching module configured to take into account the constraints associated with the data thus pulled for a particular industry type;
a processing engine configured to process the data in combination with the constraints to determine an optimized solution with respect to said constraint, the processing • engine further comprising:
an optimization module configured to form an objective function of said constraint by running a space reducing robust mechanism for reducing errors while determining the optimal solution; and an output generation module configured to generate the optimal solution with respect to the data thus captured, such that the optimal solution are used for one or more utility purpose.
22. A method to facilitate energy optimization with respect to one or more constraints for an industry type for a utility purpose, the method comprising steps of:
integrating one or more source systems distributed across a network;
pulling data from the source systems thus integrated throughout the network;
taking into account the constraints associated with the data thus pulled for a particular industry type;
processing the data in combination with the constraints to determine an optimized solution with respect to said constraint, the processing further comprising:
forming an objective function of said constraint by running a space reducing robust mechanism for reducing errors while determining the optimal solution; and generating an optimal solution with respect to the data thus captured, such that the optimal solution is used, for one or more utility purpose.
23. The method of claim 4 and 22, wherein the source system further comprises of one or more forecasting systems, PPA, Hydel generation, Thermal generation, Gas generation, Bilateral, Exchange, UI.
24. The method of claim 4 and 22, wherein the energy further includes electrical energy.
25. The method of claim 4 and 22, wherein the data pulled further comprises of data related to price and availability of each of power sources, price of deviating from the agreed volume, reliability of each power source, corridor limit and transmission constraints, demand forecast, UI forecast, exchange price forecast, technical specifications on each type of power source.
26. The method of claim 4 and 22, wherein the constraints further comprises of constraints related to how much power will be available from each source at every point in time, approximate figure of the price of power in the exchange and grid, price and volume constraint for each source, grid stability related constraints, technical constraints relate to running the power plants, reliability requirement of each node in the network etc.
27. The method claim 4 and 22, wherein the space reducing robust mechanism further comprises of a mathematical programming approach, a heuristic approach or a combination thereof.
28. The method of claim 4 and 22, wherein the optimal solution further comprises of a short term energy portfolio optimization and a medium term energy portfolio optimization.
29. The method of claim 4, 22 and 28 wherein the optimal solutions are used for utility purpose selected from a group of Prioritize across contracts, spot market and captive generation, Amount of volume to be exercised across contract, how much to buy or sell (depending on the supply Available) from day-ahead exchange and spot market, how much to generate if at all at any given point in time, given current status of various generation units and their marginal cost and the spot price, Prioritize Ramp-up & Ramp Down, which unit to shut down in case of excess capacity or a combination thereof.
PCT/IN2012/000845 2011-12-23 2012-12-24 System and method facilitating forecasting, optimization and visualization of energy data for an industry WO2013102932A2 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
IN3647MU2011 2011-12-23
IN3647/MUM/2011 2011-12-23

Publications (2)

Publication Number Publication Date
WO2013102932A2 true WO2013102932A2 (en) 2013-07-11
WO2013102932A3 WO2013102932A3 (en) 2014-06-26

Family

ID=48745511

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IN2012/000845 WO2013102932A2 (en) 2011-12-23 2012-12-24 System and method facilitating forecasting, optimization and visualization of energy data for an industry

Country Status (1)

Country Link
WO (1) WO2013102932A2 (en)

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017007942A1 (en) * 2015-07-07 2017-01-12 View, Inc. Control methods for tintable windows
USD781306S1 (en) 2015-01-27 2017-03-14 Johnson Controls Technology Company Display screen or portion thereof with graphical user interface
US9638978B2 (en) 2013-02-21 2017-05-02 View, Inc. Control method for tintable windows
US9645465B2 (en) 2011-03-16 2017-05-09 View, Inc. Controlling transitions in optically switchable devices
US9927674B2 (en) 2011-03-16 2018-03-27 View, Inc. Multipurpose controller for multistate windows
EP3197001A4 (en) * 2014-09-16 2018-05-02 Sekisui Chemical Co., Ltd. Power management system and power management method
US10048561B2 (en) 2013-02-21 2018-08-14 View, Inc. Control method for tintable windows
US20190067946A1 (en) * 2017-08-28 2019-02-28 General Electric Company System and method for distribution load forecasting in a power grid
US10254618B2 (en) 2011-10-21 2019-04-09 View, Inc. Mitigating thermal shock in tintable windows
US10504195B2 (en) 2017-11-13 2019-12-10 Mitsubishi Electric Research Laboratories, Inc. System and method for decentralized energy production
US10520784B2 (en) 2012-04-17 2019-12-31 View, Inc. Controlling transitions in optically switchable devices
WO2020014006A1 (en) * 2018-07-13 2020-01-16 Siemens Aktiengesellschaft Dynamic photovoltaic curtailment for electrical power distribution
CN110866591A (en) * 2019-10-28 2020-03-06 浙江大学 Method for carrying out prospective cloud manufacturing service lease configuration based on demand prediction
CN112926725A (en) * 2020-12-01 2021-06-08 杭州电力设备制造有限公司 Data analysis method and system based on thermal image
CN113169579A (en) * 2018-06-22 2021-07-23 莫伊沙能源控股有限公司 System for machine learning, local multi-asset flexibility optimization and management of distributed energy storage resources
US11255722B2 (en) 2015-10-06 2022-02-22 View, Inc. Infrared cloud detector systems and methods
CN114172146A (en) * 2021-11-16 2022-03-11 国网辽宁省电力有限公司 Intelligent auxiliary decision-making system for real-time visualization power grid service demand planning
EP3843229A4 (en) * 2018-08-23 2022-05-18 Hitachi, Ltd. Power transmission loss calculation device and power transmission loss calculation method
US11635666B2 (en) 2012-03-13 2023-04-25 View, Inc Methods of controlling multi-zone tintable windows
US11674843B2 (en) 2015-10-06 2023-06-13 View, Inc. Infrared cloud detector systems and methods
US11719990B2 (en) 2013-02-21 2023-08-08 View, Inc. Control method for tintable windows
CN117277439A (en) * 2023-09-26 2023-12-22 宁德时代新能源科技股份有限公司 Power grid dispatching method, device, equipment and storage medium
US11950340B2 (en) 2012-03-13 2024-04-02 View, Inc. Adjusting interior lighting based on dynamic glass tinting
US11960190B2 (en) 2019-03-20 2024-04-16 View, Inc. Control methods and systems using external 3D modeling and schedule-based computing

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003023675A2 (en) * 2001-09-13 2003-03-20 Abb Ab Method and system to calculate a demand for energy
CN101682195A (en) * 2007-03-26 2010-03-24 Vpec株式会社 Power system
CN102118123A (en) * 2010-01-05 2011-07-06 通用电气公司 Method and apparatus for monitoring and controlling a power system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003023675A2 (en) * 2001-09-13 2003-03-20 Abb Ab Method and system to calculate a demand for energy
CN101682195A (en) * 2007-03-26 2010-03-24 Vpec株式会社 Power system
CN102118123A (en) * 2010-01-05 2011-07-06 通用电气公司 Method and apparatus for monitoring and controlling a power system

Cited By (40)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11520207B2 (en) 2011-03-16 2022-12-06 View, Inc. Controlling transitions in optically switchable devices
US10908470B2 (en) 2011-03-16 2021-02-02 View, Inc. Multipurpose controller for multistate windows
US9645465B2 (en) 2011-03-16 2017-05-09 View, Inc. Controlling transitions in optically switchable devices
US9927674B2 (en) 2011-03-16 2018-03-27 View, Inc. Multipurpose controller for multistate windows
US10712627B2 (en) 2011-03-16 2020-07-14 View, Inc. Controlling transitions in optically switchable devices
US10254618B2 (en) 2011-10-21 2019-04-09 View, Inc. Mitigating thermal shock in tintable windows
US11635666B2 (en) 2012-03-13 2023-04-25 View, Inc Methods of controlling multi-zone tintable windows
US11950340B2 (en) 2012-03-13 2024-04-02 View, Inc. Adjusting interior lighting based on dynamic glass tinting
US10520784B2 (en) 2012-04-17 2019-12-31 View, Inc. Controlling transitions in optically switchable devices
US9638978B2 (en) 2013-02-21 2017-05-02 View, Inc. Control method for tintable windows
US11899331B2 (en) 2013-02-21 2024-02-13 View, Inc. Control method for tintable windows
US11940705B2 (en) 2013-02-21 2024-03-26 View, Inc. Control method for tintable windows
US10539854B2 (en) 2013-02-21 2020-01-21 View, Inc. Control method for tintable windows
US10048561B2 (en) 2013-02-21 2018-08-14 View, Inc. Control method for tintable windows
US11719990B2 (en) 2013-02-21 2023-08-08 View, Inc. Control method for tintable windows
US11126057B2 (en) 2013-02-21 2021-09-21 View, Inc. Control method for tintable windows
US10802372B2 (en) 2013-02-21 2020-10-13 View, Inc. Control method for tintable windows
EP3197001A4 (en) * 2014-09-16 2018-05-02 Sekisui Chemical Co., Ltd. Power management system and power management method
USD781306S1 (en) 2015-01-27 2017-03-14 Johnson Controls Technology Company Display screen or portion thereof with graphical user interface
US11261654B2 (en) 2015-07-07 2022-03-01 View, Inc. Control method for tintable windows
WO2017007942A1 (en) * 2015-07-07 2017-01-12 View, Inc. Control methods for tintable windows
CN107850815A (en) * 2015-07-07 2018-03-27 唯景公司 Control method for pigmentable window
TWI746446B (en) * 2015-07-07 2021-11-21 美商唯景公司 Viewcontrol methods for tintable windows
US11674843B2 (en) 2015-10-06 2023-06-13 View, Inc. Infrared cloud detector systems and methods
US11255722B2 (en) 2015-10-06 2022-02-22 View, Inc. Infrared cloud detector systems and methods
EP3451479A1 (en) * 2017-08-28 2019-03-06 General Electric Company System and method for distribution load forecasting in a power grid
US10700523B2 (en) 2017-08-28 2020-06-30 General Electric Company System and method for distribution load forecasting in a power grid
US20190067946A1 (en) * 2017-08-28 2019-02-28 General Electric Company System and method for distribution load forecasting in a power grid
US10504195B2 (en) 2017-11-13 2019-12-10 Mitsubishi Electric Research Laboratories, Inc. System and method for decentralized energy production
CN113169579A (en) * 2018-06-22 2021-07-23 莫伊沙能源控股有限公司 System for machine learning, local multi-asset flexibility optimization and management of distributed energy storage resources
WO2020014006A1 (en) * 2018-07-13 2020-01-16 Siemens Aktiengesellschaft Dynamic photovoltaic curtailment for electrical power distribution
EP3843229A4 (en) * 2018-08-23 2022-05-18 Hitachi, Ltd. Power transmission loss calculation device and power transmission loss calculation method
US11960190B2 (en) 2019-03-20 2024-04-16 View, Inc. Control methods and systems using external 3D modeling and schedule-based computing
CN110866591A (en) * 2019-10-28 2020-03-06 浙江大学 Method for carrying out prospective cloud manufacturing service lease configuration based on demand prediction
CN110866591B (en) * 2019-10-28 2022-11-01 浙江大学 Method for carrying out prospective cloud manufacturing service lease configuration based on demand prediction
CN112926725A (en) * 2020-12-01 2021-06-08 杭州电力设备制造有限公司 Data analysis method and system based on thermal image
CN114172146B (en) * 2021-11-16 2023-06-09 国网辽宁省电力有限公司 Real-time visual power grid business demand planning intelligent auxiliary decision-making system
CN114172146A (en) * 2021-11-16 2022-03-11 国网辽宁省电力有限公司 Intelligent auxiliary decision-making system for real-time visualization power grid service demand planning
CN117277439A (en) * 2023-09-26 2023-12-22 宁德时代新能源科技股份有限公司 Power grid dispatching method, device, equipment and storage medium
CN117277439B (en) * 2023-09-26 2024-03-29 宁德时代新能源科技股份有限公司 Power grid dispatching method, device, equipment and storage medium

Also Published As

Publication number Publication date
WO2013102932A3 (en) 2014-06-26

Similar Documents

Publication Publication Date Title
US11398000B2 (en) Methods and systems for machine-learning for prediction of grid carbon emissions
WO2013102932A2 (en) System and method facilitating forecasting, optimization and visualization of energy data for an industry
Botterud et al. Wind power trading under uncertainty in LMP markets
US9134353B2 (en) Comfort-driven optimization of electric grid utilization
US11909210B2 (en) Systems and methods for load management
Vieira et al. A multiple criteria utility-based approach for unit commitment with wind power and pumped storage hydro
Finnah et al. Optimizing trading decisions of wind power plants with hybrid energy storage systems using backwards approximate dynamic programming
Geiger Strategic power plant investment planning under fuel and carbon price uncertainty
Motta et al. Survey of optimization models for power system operation and expansion planning with demand response
Binyet et al. Potential of demand response for power reallocation, a literature review
Henni et al. Industrial peak shaving with battery storage using a probabilistic forecasting approach: Economic evaluation of risk attitude
Nandakumar et al. Impact of increased renewables on natural gas markets in eastern United States
Keles et al. Evaluation of photovoltaic storage systems on energy markets under uncertainty using stochastic dynamic programming
Rathnayaka Development of a community-based framework to manage prosumers in smart grid
Bublitz Capacity remuneration mechanisms for electricity markets in transition
Skantze et al. Modelling locational price spreads in competitive electricity markets; applications for transmission rights valuation and replication
Valtonen Distributed energy resources in an electricity retailer’s short-term profit optimization
HEYMANS Modelling of Electricity Load Data for Belgium
Carvallo et al. A Guide for Improved Resource Adequacy Assessments in Evolving Power Systems
Custers Impacts of network tariff re-design within industrial parks
Ngondya Demand-side management framework for deregulated electricity markets
Tyagi et al. Transforming electrical load from an operational constraint to a controllable resource
Domínguez-Garabitos et al. Demand-Shifting Strategies to Optimize the Performance of the Wholesale Electricity Market: A Dominican Republic Study Case
Boscán F Essays on the Design of Contracts and Markets for Power System Flexibility
Helmberg et al. Finance, Regulations, Politics and Market Design

Legal Events

Date Code Title Description
122 Ep: pct application non-entry in european phase

Ref document number: 12864251

Country of ref document: EP

Kind code of ref document: A2