Claims:A method (300) for generating bids in an energy market, comprising:
generating (302) a plurality of initial supply bids for a plurality of time slots, for a seller, via one or more hardware processors;
forecasting (304) a competition bid curve using historical bidding information, via the via one or more hardware processors;
determining (306) an expected market clearing price and a cleared quantity for the seller, for each of the plurality of time slots, by applying a market clearing model on the plurality of initial supply bids and the competition bid curve, via the one or more hardware processors;
verifying (308) if the expected market clearing price and the cleared quantity satisfy a plurality of constraints, via the one or more hardware processors; and
refining (310) one or more of the plurality of initial supply bids if the expected market clearing price and cleared quantity have been verified as violating one or more of the plurality of constraints, via the one or more hardware processors.

The method as claimed in claim 1, wherein forecasting the competition bid curve comprises:
creating (402) a common price vector by taking union of price bands across various time blocks, from the historical bidding information;
mapping (404) each of a plurality of historical bids in the historical bidding information to a corresponding price band in the common price vector;
forecasting (406) bid quantity for a selected time slot at a future instance, by taking moving average estimate of bid quantities in same time slot over a plurality of past instances, wherein the forecasting bid quantity comprises an aggregate market demand and supply information;
generating (408) the competition bid curve using the forecasted bid quantity for a plurality of time slots; and
normalizing (410) the generated competition bid curve, wherein the normalized competition bid curve forms the forecasted bid curve.

The method as claimed in claim 1, wherein the market clearing model determines the expected market clearing price and the cleared quantity subjects by subjecting bid price of a plurality of buyers (buyer’s bid price), bid quantity of the plurality of buyers (buyer’s bid quantity), bid price of a plurality of sellers (seller’s bid price), and bid quantity of the plurality of sellers (seller’s bid quantity), to a condition defined as:
?maximize:??? ? {?_B¦?_n¦?(??Quantity ?_(B,n)* Price ?_(B,n) )- ? ?_S¦?_m¦?(??Quantity ?_(S,m)* Price ?_(S,m) ) ?}

where, S represents total number of sellers, B represents total number of buyers, n represents number of bids by a buyer, and m represents number of bids by a seller.

The method as claimed in claim 1, wherein verifying if the expected market clearing price and cleared quantity satisfy the plurality of constraints comprises comparing values of a plurality of slack variables associated with the cleared quantity with corresponding optimum values, wherein the values of the slack variables, if matching the corresponding optimum values, result in the generated market clearing price and the cleared quantity satisfying the plurality of constraints.

The method as claimed in claim 4, wherein refining one or more of the plurality of initial supply bids comprises adjusting value of one or more of the plurality of slack variables such that values of the plurality of slack variables match the corresponding optimum values.

A system (100) for generating bids in an energy market, comprising:
one or more hardware processors (102);
a communication interface (106); and
a memory (104) storing a plurality of instructions, wherein the plurality of instructions when executed, cause the one or more hardware processors to:
generate a plurality of initial supply bids for a plurality of time slots, for a seller;
forecast a competition bid curve using historical bidding information;
determine an expected market clearing price and a cleared quantity for the seller, for each of the plurality of time slots, by applying a market clearing model on the plurality of initial supply bids and the competition bid curve;
verify if the expected market clearing price and the cleared quantity satisfy a plurality of constraints; and
refine one or more of the plurality of initial supply bids If the expected market clearing price and cleared quantity have been verified as violating one or more of the plurality of constraints.

The system as claimed in claim 6, wherein the system forecasts the competition bid curve by:
creating a common price vector by taking union of price bands across various time blocks, from the historical bidding information;
mapping each of a plurality of historical bids in the historical bidding information to a corresponding price band in the common price vector;
forecasting bid quantity for a selected time slot at a future instance, by taking moving average estimate of bid quantities in same time slot over a plurality of past instances, wherein the forecasting bid quantity comprises an aggregate market demand and supply information;
generating the competition bid curve using the forecasted bid quantity for a plurality of time slots; and
normalizing the generated competition bid curve, wherein the normalized competition bid curve forms the forecasted bid curve.

The system as claimed in claim 6, wherein the system determines, using the market clearing model, the expected market clearing price and the cleared quantity by subjecting bid price of a plurality of buyers (buyer’s bid price), bid quantity of the plurality of buyers (buyer’s bid quantity), bid price of a plurality of sellers (seller’s bid price), and bid quantity of the plurality of sellers (seller’s bid quantity), to a condition defined as:
?maximize:??? ? {?_B¦?_n¦?(??Quantity ?_(B,n)* Price ?_(B,n) )- ? ?_S¦?_m¦?(??Quantity ?_(S,m)* Price ?_(S,m) ) ?}

where, S represents total number of sellers, B represents total number of buyers, n represents number of bids by a buyer, and m represents number of bids by a seller.

The system as claimed in claim 6, wherein the system verifies if the expected market clearing price and cleared quantity satisfy the plurality of constraints by comparing values of a plurality of slack variables associated with the cleared quantity with corresponding optimum values, wherein the values of the slack variables, if matching the corresponding optimum values, result in the generated market clearing price and the cleared quantity satisfying the plurality of constraints.

The system as claimed in claim 9, wherein the system refines one or more of the plurality of initial supply bids by adjusting value of one or more of the plurality of slack variables such that values of the plurality of slack variables match the corresponding optimum values.
, Description:FORM 2

THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003

COMPLETE SPECIFICATION
(See Section 10 and Rule 13)

Title of invention:
METHOD AND SYSTEM FOR BID FORECASTING

Applicant:
Tata Consultancy Services Limited
A company Incorporated in India under the Companies Act, 1956
Having address:
Nirmal Building, 9th Floor,
Nariman Point, Mumbai 400021,
Maharashtra, India

The following specification particularly describes the invention and the manner in which it is to be performed.
TECHNICAL FIELD
The disclosure herein generally relates to energy markets, and, more particularly, to a method and system for bid forecasting.

BACKGROUND
Sustainability and reliability are two important aspects of smart grids. Sustainability refers to higher utilization of green energy. However, these green energy sources e.g., solar PV or wind are intermittent due to weather dependency, which in turn gives rise to reliability concerns. Though electrical generation across the world is transforming from fossil fuel power plants to greener plants, a complete transformation has not happened yet. It may take few more years to achieve sustained power generation with zero carbon emissions. Conventional fossil fuel-based generators continue to operate until the reliability concerns of renewable generation are adequately addressed.
Fossil fuel generators participate in electricity markets alongside renewable plants to sell their generation. Renewable power plants have low operational costs and hence can sell their energy at a low price. On the other hand, fossil fuel power plants have considerable operational costs due to factors such as fuel and maintenance. Moreover, these generators may not be able to sell their energy consistently at high prices (to recover their operating cost with some profit margins) since the renewable plants may undercut their bids in the electricity market. Consequently, the profit margins of the conventional generators shrink with increasing renewable penetration. Further, the conventional generators are beset by constraints in their operational characteristics such as limited ramp up/down rates, minimum capacity limit at run time, emission constraints, and so on.
To remain competitive in the supply eco-system, these fossil fuel generators need a bidding strategy for participating in electricity markets that not only reflects their production costs but also their operational constraints. Some systems that are capable of generating bidding strategy predictions exist. However, bidding strategy generation and bid prediction, if not done taking into consideration certain operational constraints, would affect accuracy and usefulness of the bids generated. The state of the art systems fail to generate the bids satisfying the operational constraints. As a result, the generated bids and bidding strategies that are generated may be agnostic to the operating constraints, and this in turn reduces the profit margin of the generators. Also, the constraint violations may prevent the generator from dispatching the market cleared amount, leading to financial penalties.

SUMMARY
Embodiments of the present disclosure present technological improvements as solutions to one or more of the above-mentioned technical problems recognized by the inventors in conventional systems. For example, in one embodiment, a method for generating bids in an energy market is provided. In this method, initially a plurality of initial supply bids are generated for a plurality of time slots, for a seller, via one or more hardware processors. Further, a competition bid curve is forecasted using historical bidding information, via the via one or more hardware processors. Further, an expected market clearing price and a cleared quantity for the seller are determined, for each of the plurality of time slots, by applying a market clearing model on the plurality of initial supply bids and the competition bid curve, via the one or more hardware processors. Further, it is verified if the expected market clearing price and the cleared quantity satisfy a plurality of constraints, via the one or more hardware processors. If the expected market clearing price and cleared quantity have been verified as violating one or more of the plurality of constraints, then one or more of the plurality of initial supply bids are refined via the one or more hardware processors.
In another aspect, a method of forecasting the competition bid curve is provided. The method involves the following steps. Initially, a common price vector is created by taking union of price bands across various time blocks, from the historical bidding information. Further, each of a plurality of historical bids in the historical bidding information is mapped to a corresponding price band in the common price vector. Further, a bid quantity for a selected time slot at a future instance is predicted by taking moving average estimate of bid quantities in same time slot over a plurality of past instances, wherein the forecasting bid quantity comprises an aggregate market demand and supply information. Further, the competition bid curve is generated using the forecasted bid quantity for a plurality of time slots. Further, the generated competition bid curve is normalized, and the normalized competition bid curve forms the forecasted bid curve.
In yet another aspect, a system for generating bids in an energy market is provided. The system includes one or more hardware processors, a communication interface, and a memory storing a plurality of instructions. The plurality of instructions when executed, cause the one or more hardware processors to initially generate a plurality of initial supply bids for a plurality of time slots, for a seller. The system then forecasts a competition bid curve using historical bidding information. The system then determines an expected market clearing price and a cleared quantity for the seller, for each of the plurality of time slots, by applying a market clearing model on the plurality of initial supply bids and the competition bid curve. The system then verifies if the expected market clearing price and the cleared quantity satisfy a plurality of constraints. If the expected market clearing price and cleared quantity have been verified as violating one or more of the plurality of constraints, then the one or more of the plurality of initial supply bids are refined.
In yet another aspect, a system for forecasting the competition bid curve is provided. The system creates a common price vector by taking union of price bands across various time blocks, from the historical bidding information. The system then maps each of a plurality of historical bids in the historical bidding information to a corresponding price band in the common price vector. The system then forecasts bid quantity for a selected time slot at a future instance, by taking moving average estimate of bid quantities in same time slot over a plurality of past instances, wherein the forecasting bid quantity comprises an aggregate market demand and supply information. Further the system generates the competition bid curve using the forecasted bid quantity for a plurality of time slots, and then the generated competition bid curve is normalized.
In yet another aspect, a non-transitory computer readable medium for generating bids in an energy market is provided. The non-transitory computer readable medium includes a plurality of instructions, which when executes, cause the following steps. Initially a plurality of initial supply bids are generated for a plurality of time slots, for a seller, via one or more hardware processors. Further, a competition bid curve is forecasted using historical bidding information, via the via one or more hardware processors. Further, an expected market clearing price and a cleared quantity for the seller are determined, for each of the plurality of time slots, by applying a market clearing model on the plurality of initial supply bids and the competition bid curve, via the one or more hardware processors. Further, it is verified if the expected market clearing price and the cleared quantity satisfy a plurality of constraints, via the one or more hardware processors. If the expected market clearing price and cleared quantity have been verified as violating one or more of the plurality of constraints, then one or more of the plurality of initial supply bids are refined via the one or more hardware processors.
In yet another aspect, a non-transitory computer readable medium for forecasting the competition bid curve is provided. The non-transitory computer readable medium includes a plurality of instructions, which when executes, cause the following steps. Initially, a common price vector is created by taking union of price bands across various time blocks, from the historical bidding information. Further, each of a plurality of historical bids in the historical bidding information is mapped to a corresponding price band in the common price vector. Further, a bid quantity for a selected time slot at a future instance is predicted by taking moving average estimate of bid quantities in same time slot over a plurality of past instances, wherein the forecasting bid quantity comprises an aggregate market demand and supply information. Further, the competition bid curve is generated using the forecasted bid quantity for a plurality of time slots. Further, the generated competition bid curve is normalized.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

BRIEF DESCRIPTION OF THE DRAWINGS
The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, serve to explain the disclosed principles:
FIG. 1 illustrates an exemplary system for bid generation, according to some embodiments of the present disclosure.
FIG. 2 is a functional block diagram of the system of FIG. 1, according to some embodiments of the present disclosure.
FIG. 3 illustrates a flow diagram depicting steps involved in the process of generating bid predictions by the system of FIG. 1, in accordance with some embodiments of the present disclosure.
FIG. 4 illustrates a flow diagram depicting steps involved in the process of forecasting competition `bid curve, by the system of FIG. 1, in accordance with some embodiments of the present disclosure.
FIGS. 5A through 5D are example graphs depicting values of different parameters as obtained during experimental bidding predictions conducted using the system of FIG. 1 and a plurality of state of the art approaches, in accordance with some embodiments of the present disclosure.

DETAILED DESCRIPTION OF EMBODIMENTS
Exemplary embodiments are described with reference to the accompanying drawings. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. Wherever convenient, the same reference numbers are used throughout the drawings to refer to the same or like parts. While examples and features of disclosed principles are described herein, modifications, adaptations, and other implementations are possible without departing from the scope of the disclosed embodiments.
Referring now to the drawings, and more particularly to FIG. 1 through FIG. 5D, where similar reference characters denote corresponding features consistently throughout the figures, there are shown preferred embodiments and these embodiments are described in the context of the following exemplary system and/or method.
FIG. 1 illustrates an exemplary system for bid generation, according to some embodiments of the present disclosure. In an embodiment, the system 100 includes a processor(s) 102, communication interface device(s), alternatively referred as input/output (I/O) interface(s) 106, and one or more data storage devices or a memory 104 operatively coupled to the processor(s) 102. A functional implementation of the system 100 is given in FIG. 2. The components of the system 100 as depicted in FIG. 2 may be implementation of the one or more hardware processors, and are configured to execute various actions associated with the bid generation.
Referring to the components of system 100, in an embodiment, the processor(s) 102, can be one or more hardware processors 102. In an embodiment, the one or more hardware processors 102 can be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions. Among other capabilities, the one or more hardware processors 102 are configured to fetch and execute computer-readable instructions stored in the memory 104. In an embodiment, the system 100 can be implemented in a variety of computing systems including laptop computers, notebooks, hand-held devices such as mobile phones, workstations, mainframe computers, servers, and the like.
The I/O interface(s) 106 can include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface to display the generated target images and the like and can facilitate multiple communications within a wide variety of networks N/W and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular and the like. In an embodiment, the I/O interface (s) 106 can include one or more ports for connecting to a number of external devices or to another server or devices.
The memory 104 may include any computer-readable medium known in the art including, for example, volatile memory, such as static random access memory (SRAM) and dynamic random access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes.
Further, the memory 104 includes a database 108 that stores all data associated with the bid forecasting, being performed by the system 100. For example, the database 108 stores the configurable instructions that are executed to cause the one or more hardware processors 102 to perform various steps associated with the blast furnace optimization. The database 108 may further store all data, i.e. generated during the bid forecasting being done by the system 100, for example, i.e. data such as but not limited to the historical bidding information collected, bids generated, determines cleared quantity and price, and so on.
The functional implementation of the system 100, as depicted in FIG. 2 includes a bid forecasting module 201, a market simulation module 202, a bid forecasting module 203, a bid refinement module 204, and a constraint satisfaction module 205. Functions of the components of the system 100 as in FIG. 1 and FIG. 2 are explained with reference to explanation of the steps depicted in FIG. 3 and FIG. 4.
FIG. 3 illustrates a flow diagram depicting steps involved in the process of generating bid predictions by the system of FIG. 1, in accordance with some embodiments of the present disclosure. The method depicted in FIG. 3 is denoted as method 300, with each step being represented by different numerals starting from 302. Initially, at step 302, the system 100 generates a plurality of initial supply bids F ^_(s,h)for a generator (or seller) s for whom the bid generation is being performed, based on marginal production costs incurred. The system 100 may use any suitable technique such as but not limited to a) bidding at cost of generation, b) bidding at previous day market price, and any other strategic approach, to generate the initial supply bids. The approaches used by these techniques may be different. For example, the initial supply bids are generated using a strategy that focuses on profit maximization. In an embodiment, the initial supply bids are obtained as inputs by the system 100.
At step 304, the system 100 forecasts, using the bid curve forecasting module 201, a competition bid curve by processing a historical bidding information. The historical bidding information includes data such as but not limited to bids generated, market clearing price, and cleared quantity, at a plurality of past instances of bidding. The historical bidding information may be pre-stored in a database in the memory 104, or may be dynamically accessed from any external database via suitable interfaces/channels provided by the I/O interface 106. In another embodiment, the historical bidding information stored in the database in the memory 104 is updated periodically. In an embodiment, the steps 302 and 304 may be executed in parallel Various steps involved in the process of forecasting the competition bid curve are depicted by the steps of method 400 in FIG. 4, and are explained hereafter.
The bid curve forecasting involves forecasting a) an aggregated supply bid curve (from among a set of sellers, after considering operational cost incurred while generating the energy, represented as S ¯), and an aggregated demand bid curve (from a set of buyers B) of all time slots on a day for which the prediction is being generated. At step 402, the system 100 creates a common price vector by taking union of a plurality of price bands across various time blocks in the historical bidding information. The ‘price bands’ may indicate different price segments in which bidding of different quantities took place. The historical bidding information may contain information in the form of price-quantity pairs, for different time slots over a period of time. Further, at step 404, the system 100 maps each of the plurality of historical bids in the historical bidding information to a corresponding price band in the common price vector.
Further, at step 406, the system 100 forecasts bid quantity for a selected time slot at a future instance, using the bid forecasting module 203, by taking moving average estimate of bid quantities in same time slot over a plurality of past instances. The forecasted bid quantity contains the aggregate market demand and supply information. Further, at step 408, the system 100 generates a competition bid curve using the forecasted bid quantity, for a plurality of time slots. Then at step 410, the system 100 normalizes the generated competition bid curve. By means of normalization, the system 100 reduces effect of exogeneous disturbances such as weather. In order to perform the normalization, the system 100 may forecast an aggregate demand and supply information for each hour h, using a multi-layer neural network and may then use the forecasted aggregate demand and supply information for normalizing the demand and supply bid curves.
After forecasting the competition bid curve at step 306, further at step 308, the system 100 determines using the market simulation module 202, an expected market clearing price and a cleared quantity for the seller, for each of a plurality of time slots, by applying a market clearing model on the initial supply bids and the competition bid curve. In a day ahead double sided electricity market in which the buyers as well as sellers submit price and quantity bids for each time slot. Each buyer b may submit n bids in the form of price-quantity pairs, and the seller submits m bids. The market clearing model determines an expected market clearing price and the cleared quantity by subjecting bid price of a plurality of buyers (buyer’s bid price), bid quantity of the plurality of buyers (buyer’s bid quantity), bid price of a plurality of sellers (seller’s bid price), and bid quantity of the plurality of sellers (seller’s bid quantity), to a condition defined as:
?maximize:??? ? {?_B¦?_n¦?(??Quantity ?_(B,n)* Price ?_(B,n) )- ? ?_S¦?_m¦?(??Quantity ?_(S,m)* Price ?_(S,m) ) ?}

--- (1)
where, S represents total number of sellers, B represents total number of buyers, n represents number of bids by a buyer, and m represents number of bids by a seller.
The bid forecasting module 203 of the system 100 determines an expected market clearing price and the cleared quantity for the seller, using the market clearing model, such that the determined expected market clearing price and the cleared quantity satisfy the condition in (1).
In any energy market, the generator (who may also be the seller) has multiple operational constraints (also referred to as ‘constraints’). Some examples of the constraints are, but not limited to limits on how quickly the generation can be ramped up/down across time slots, the minimum quantity that needs to be output when the generation is on, and the least amount of time before which a generator cannot be turned on after it has been turned off. The constraint satisfaction module 205 processes the determined expected market clearing price and the cleared quantity to verify if they satisfy one of more of the constraints which are pre-configured with the constraint satisfaction module 205, based on business and operational requirements of the generator.
For explanation purpose, working of the constraint satisfaction module 205 is explained by considering two constraints i.e. a) ramping limit constraints, and b) minimum generation limit constraints. It is to be noted that the explanation of working of the constraint satisfaction module 205 by considering only these two constraints is not intended to limit the scope in any manner, and the constraint satisfaction module 205 can be configured to function with any number of constraints as may be required by different implementations.
The system 100 determines, using the equations 2 through 5, an optimal quantum of change needed in the cleared generation volume across time slots. In each time slot h, it is possible that the cleared quantity C_(s,h) violates one or more of the constraints i.e. the ramping limit constraints, and the minimum generation limit constraints independently. The system 100 initially verifies if the expected market clearing price and the cleared quantity satisfy the plurality of constraints, by comparing values of a plurality of slack variables associated with the cleared quantity with corresponding optimum values, wherein the values of the slack variables, if matching the corresponding optimum values, result in the generated market clearing price and the cleared quantity satisfying the plurality of constraints. For example, to verify if the expected market clearing price and the cleared quantity satisfy the constraints being considered (i.e. the minimum generation limit, max ramp-up and max ramp-down), the system 100 uses three slack variables slack variable for minimum generation condition (?sm?_(h)), slack variable for ramp-up condition (?su?_h), and slack variable for ramp-down condition (sd_h).
Comparison of value of each of these slack variables, calculated using equations (2) to (5), with the corresponding optimum values also helps the system 100 determine to what extent the determined generated quantity is to be changed so that the cleared quantity satisfies the constraints. In an embodiment, as value of the generated quantity may have been selected to meet one or other business requirements, by a plant operator (or generator), the system 100 is configured to make minimum changes to the determined generated quantity if one or more of the constraints are not satisfied, wherein the term ‘minimum change’ is interpreted as a change that would result in the bids satisfying all of the constraints. To meet this requirement, an objective for changing the determined generated quantity is defined as:
¦(max@sm,su,sd)?_h¦?P_h^* (C_(s,h)-? sm?_h-? su?_h- ? sd?_h ) ? --- (2)

Subject to
C_(s,min)= C_(s,h) + ? sm?_h = C_(s,max),?h --- (3)

(C_(s,h)+? sm?_h) - (C_(s,h-1)-? sm?_(h-1)) - ?su?_h=R_max^up,?h --- (4)

(C_(s,h-1)+? sm?_(h-1)) - (C_(s,h)-? sm?_h) - ?sd?_h=R_max^down,?h --- (5)

Equation (3) is used by the system 100 to determine value of the slack variable ? sm?_h by which the cleared quantity is to be boosted at slot h to satisfy the minimum generation constraint. Equations (4) and (5) are used to determine values of ? su?_h and ? sv?_h respectively, by which the cleared quantity at h is to be increased/decreased to satisfy the ramping up/down constraints.
Further at step 310, the bid refinement module 204 refines the bids, based on the difference in values of the slack variables obtained at step 308, such that the cleared quantity satisfies the constraints. In an embodiment, the bid refinement module 204 performs an iterative process for the purpose of refining the bids, till the determined values of the slack variables match the corresponding optimum values.
In various embodiments, the steps in method 300 may be performed in the same order as depicted in FIG. 3 or in any alternate order that is technically feasible. In another embodiment, one or more steps in method 300 may be omitted.

Experimental Results:-
During the experiments conducted, in order to test the performance of the bid generation approach, a single thermal power generating unit participating in the double-sided, day-ahead electricity market of European Power Exchange (EPEX) was considered. Market logs from EPEX, containing historical logs of market clearing prices, demand and supply bids was used. Specifically, hourly information of demand and supply bids with multiple price and quantity pairs for France was used. The bid prices ranged from 500 ¢/MWh to 3000 ¢/MWh. The characteristics of the thermal generator considered are given in Table I. While a thermal generator with a single unit was considered during the experiments, the method 300 can be appropriately extended to generation company owning multiple generating units with different constraints and parameters. The optimization model was formulated using Pyomo with GNU Linear Programming Kit (GLPK) solver.
Parameter Value
Min Gen
Max Gen
Max Ramp Up Limit
Max Ramp Down Limit 150 MW
400 MW
120 MW/h
120 MW/h
Table. 1
Performance of bid curve forecasting

Accuracy of the bid curve forecasting technique was tested using historical market logs of EPEX day ahead market. The forecasting accuracy was tested with respect to both buy bids and sell bids. These curves are estimated for each hour of the test period i.e. in this example, 01-Feb-2019 to 28- Feb-2019. The overall market supply and demand values for the test period was forecasted using Long Short-Term Memory (LSTM) networks. FIG. 5A shows the forecasted aggregated supply and demand bid curves for a sample hour.
FIG. 5B shows the performance of our bid curve forecasting technique in terms of the Mean Absolute Percentage Error (MAPE) for supply and demand bids. Medians of hourly MAPE for supply and demand bids are 10% and 13% respectively for Feb 2019. FIG. 5C shows the daily MAPE for market clearing price (MCP) derived from the forecasted supply and demand bid curves. Median of the MAPE for MCP is 11%. Median was taken to remove the influence of outliers on the average for Feb 2019.
Performance of the bidding approach in method 300
The performance of the bidding generation approach was tested using the day-ahead market logs obtained from EPEX. Performance of the following three approaches (two baselines and proposed) were compared.
Zero Price approach: In this approach, the generator places its entire generation at a price of zero.
Marginal Cost approach: Here, the generator places its bids at the price bands corresponding to the marginal cost of generation.
Method 300: Here bids are placed at marginal cost of generation followed by bid refinement.
Table. 2 provides a summary of profit and volume statistics for a simulation duration of one month. It can be noticed that maximum volume is cleared when all the quantity is bid at zero price. In this case, the generator is acting only as a price taker and is missing the opportunity to earn better profits as compared to other cases. Under the marginal cost approach, the generator aims to recover its cost of generation and tries to exploit the possibility of being a price maker. It takes the risk of volume getting uncleared. Further, the volume cleared across time slots violates the technical constraints of the generator. If the generator would not be able to sell this in some other markets, it would be result in more loss than the bidding at zero price.

Zero price Marginal cost Proposed
Volume cleared (MW) 241,652 167, 456 184, 179
Volume Violated (MW) 0 34, 430 18, 677
Revenue from volume cleared (Million ¢) 10,935 8.395 8.955
Cost of volume cleared (Million ¢) 9.666 6.698 7.367
Profit (Million ¢) 1.269 1.697 1.588

Table. 2

In the approach in method 300, the generator refined its bids by simulating the competition and market in order to minimize the constraint violations. However, there may still be some violations in the volume due to error in the forecasted competition and demand bid curves when the forecasted competition bid curves be replaced by the actual for the historical period. Nonetheless, it was found that the approach in method 300 provided a reasonable trade-off for higher profit from sale and lower volume violated (FIG. 5D). It was also found that method 300 decreases the constraint violations by as much as 45% while marginally affecting the profits by just 6%. Convergence of the iterative approach in method 300 is now discussed. Across the 28 days that the test was conducted for, system 100 took converges in less than 10 iterations in most cases. However, in a few instances, the number of iterations went upto as high as 22. Higher number of iterations are usually in the case when the market clearing price is equal to the bid price and the quantity can be partially cleared at the MCP.
The written description describes the subject matter herein to enable any person skilled in the art to make and use the embodiments. The scope of the subject matter embodiments is defined by the claims and may include other modifications that occur to those skilled in the art. Such other modifications are intended to be within the scope of the claims if they have similar elements that do not differ from the literal language of the claims or if they include equivalent elements with insubstantial differences from the literal language of the claims.
The embodiments of present disclosure herein address unresolved problem of bid generation in energy markets, satisfying various operational constraints. The embodiment, thus provides a mechanism to verify if cleared quantity in an energy market satisfies various operational constraints. Moreover, the embodiments herein further provide a mechanism to automatically refine bids so as to satisfy the operational constraints.
It is to be understood that the scope of the protection is extended to such a program and in addition to a computer-readable means having a message therein; such computer-readable storage means contain program-code means for implementation of one or more steps of the method, when the program runs on a server or mobile device or any suitable programmable device. The hardware device can be any kind of device which can be programmed including e.g., any kind of computer like a server or a personal computer, or the like, or any combination thereof. The device may also include means which could be e.g., hardware means like e.g., an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a combination of hardware and software means, e.g., an ASIC and an FPGA, or at least one microprocessor and at least one memory with software processing components located therein. Thus, the means can include both hardware means and software means. The method embodiments described herein could be implemented in hardware and software. The device may also include software means. Alternatively, the embodiments may be implemented on different hardware devices, e.g., using a plurality of CPUs.
The embodiments herein can comprise hardware and software elements. The embodiments that are implemented in software include but are not limited to, firmware, resident software, microcode, etc. The functions performed by various components described herein may be implemented in other components or combinations of other components. For the purposes of this description, a computer-usable or computer readable medium can be any apparatus that can comprise, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed. Alternatives (including equivalents, extensions, variations, deviations, etc., of those described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope of the disclosed embodiments. Also, the words “comprising,” “having,” “containing,” and “including,” and other similar forms 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.
Furthermore, one or more computer-readable storage media may be utilized in implementing embodiments consistent with the present disclosure. A computer-readable storage medium refers to any type of physical memory on which information or data readable by a processor may be stored. Thus, a computer-readable storage medium may store instructions for execution by one or more processors, including instructions for causing the processor(s) to perform steps or stages consistent with the embodiments described herein. The term “computer-readable medium” should be understood to include tangible items and exclude carrier waves and transient signals, i.e., be non-transitory. Examples include random access memory (RAM), read-only memory (ROM), volatile memory, nonvolatile memory, hard drives, CD ROMs, DVDs, flash drives, disks, and any other known physical storage media.
It is intended that the disclosure and examples be considered as exemplary only, with a true scope of disclosed embodiments being indicated by the following claims.