Power trading simulation device, power trading simulation method, and program

Abstract

To simulate a power transaction with high accuracy even if actual data of the power transaction are insufficient.SOLUTION: A power transaction simulation device of this embodiment comprises: a storage part for storing a plurality of business operator models for simulating a sales transaction of a power commodity at each of time zones which are obtained by dividing one day into plurality, and transaction models; and a processing part for operating a plurality of order behavior agents who decide and output sales kinds, order prices and order quantities of a power transaction on the basis of inputs of a power demand assumption, a market transaction price and a transaction contract actual result in correspondence with the plurality of business operator models, and operating transition agents who decide and output the market transaction price and the transaction contract actual results of the plurality of order behavior agents on the basis of inputs of the sales kinds, the order prices and the order quantities outputted by the plurality of order behavior agents in correspondence with the transaction models.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] An embodiment of the present invention relates to an energy trading simulation device, an energy trading simulation method, and a program. [Background technology]

[0002] In the domestic wholesale electricity market, there is the spot market, where contracts are made using the itayose method, and the hourly market for continuous trading (hereinafter simply referred to as "continuous trading"). In the spot market, contracts are made at equilibrium pricing, so the order price and the contract price do not match, and contracts may not be made as planned.

[0003] On the other hand, the Zaraba pre-hour market allows transactions to be made according to market prices up until the day of power supply, and is expected to play a role as a place to compensate for fluctuations in supply and demand plans caused by changes in forecasts of power generation, etc. In addition, with the expansion of power sources using renewable energy, transactions in the Zaraba pre-hour market are expected to increase.

[0004] Methods for supporting such continuous electricity trading include, for example, a method for simulating market price fluctuations using actual market data, and a method for predicting electricity prices based on environmental data and actual contract volume data.

[0005] It is also expected that P2P (Peer to Peer) electricity trading will take place via electricity auctions between surplus electricity from consumers who own power generation facilities that use renewable energy sources such as solar power generation, and consumers who want to purchase electricity generated from renewable energies at low cost. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2020-184246 A [Patent Document 2] JP 2019-46281 A Summary of the Invention [Problem to be solved by the invention]

[0007] In addition, the market price of electricity changes over time depending on multiple factors such as the balance between electricity supply and demand, and the investment behavior of consumers. However, in the conventional technology, it was not possible to simulate electricity trading with a high degree of accuracy when there was insufficient data on the actual electricity trading.

[0008] In order to achieve the above object, an embodiment of the present invention aims to provide an energy trading simulation device, an energy trading simulation method, and a program that are capable of simulating energy trading with high accuracy even when actual energy trading data is insufficient. [Means for solving the problem]

[0009] The energy trading simulator of the embodiment includes a memory unit that stores a plurality of operator models for simulating buying and selling transactions of energy products for each time period into which a day is divided, and a processing unit that operates a plurality of order behavior agents that determine and output a buying and selling type, order price, and order quantity of an energy trading based on inputs of an electricity supply and demand forecast, a market trading price, and trading agreement results, corresponding to each of the plurality of operator models, and that determines and output the market trading price and the trading agreement results of each of the plurality of order behavior agents, corresponding to the trading model, based on inputs of the buying and selling type, the order price, and the order quantity output by each of the plurality of order behavior agents. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic diagram showing an overview of simulation of energy trading in the first embodiment. [Diagram 2] FIG. 2 is a diagram illustrating a functional configuration of the information processing device in the first embodiment. [Diagram 3] FIG. 3 is a flowchart showing the process of simulating energy trading by the information processing device of the first embodiment. [Figure 4] FIG. 4 is a diagram illustrating the cost characteristics and utility characteristics of the consumer model in the first embodiment. [Diagram 5] FIG. 5 is an explanatory diagram of a method of calculating a base price for continuous orders according to the execution volume performance in the first embodiment. [Figure 6] FIG. 6 is a diagram showing the relationship between continuous order characteristics and order prices in the first embodiment. [Figure 7] FIG. 7 is an explanatory diagram of the continuous auction model in the first embodiment. [Figure 8] FIG. 8 is a diagram showing an example of a simulation result of a continuous auction in the first embodiment. [Figure 9] FIG. 9 is a schematic diagram showing an overview of the simulation of energy trading in the second embodiment. [Figure 10] FIG. 10 is an explanatory diagram of the cost characteristics of the power generation company model and the utility characteristics of the retail company model in the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, embodiments (first and second embodiments) of the energy trading simulator, the energy trading simulation method, and the program of the present invention will be described with reference to the drawings. As described above, the electricity market is divided into a spot market where contracts are made by the book-trading method, and a continuous aforesaid market. In the spot market, power generation companies and retail companies make bids according to their businesses in the energy trading market, so that all bids from power generation companies and retail companies are compared on a daily basis, and price and quantity are balanced based on the relationship between supply and demand.

[0012] In addition, the actual supply-demand relationship is not fixed, and there are fluctuations in the power supply due to breakdowns in power generation facilities, and fluctuations in power demand due to temperature changes, etc. Therefore, the pre-hour market is a place to absorb and adjust these fluctuations in the supply-demand relationship, and the price and quantity are individually agreed upon according to buy and sell orders for electricity products in 48 trading time slots, which divide a day into electricity measurement units (for example, 0 minutes to 30 minutes, 30 minutes to 60 minutes of each hour).

[0013] In the following, a first embodiment will be described with respect to a technique for simulating a P2P electricity auction transaction between consumers, and a second embodiment will be described with respect to a technique for simulating an electricity market transaction between a power generation company and a retail company.

[0014] (First embodiment) In the first embodiment, a power trading simulation will be described that targets a P2P power auction transaction between a consumer having surplus power derived from renewable energy (renewable energy) and a consumer who wants to purchase renewable energy power at low cost.

[0015] Fig. 1 is a schematic diagram showing an overview of the simulation of energy trading in the first embodiment, and Fig. 2 is a diagram showing the functional configuration of an information processing device 1 in the first embodiment.

[0016] The energy trading simulation method of the first embodiment can be realized by an information processing device 1 (energy trading simulator) shown in Fig. 2. The information processing device 1 includes a storage unit 2, an input unit 3, a display unit 4, a communication unit 5, and a processing unit 6.

[0017] The storage unit 2 is a means for storing various information, and is realized by, for example, a random access memory (RAM), a read only memory (ROM), a hard disk drive (HDD), a solid state drive (SSD), etc. Details of the storage unit 2 will be described later.

[0018] The input unit 3 is an information input means by a user, and is realized by, for example, a mouse, a keyboard, a touch panel, or the like.

[0019] The display unit 4 is an information display means, and is realized by, for example, an LCD (Liquid Crystal Display) or the like.

[0020] The communication unit 5 is a communication interface for communicating with an external device.

[0021] The processing unit 6 is a means for performing various types of arithmetic processing, and is realized by, for example, a CPU (Central Processing Unit). The processing unit 6 will be described in detail later.

[0022] The energy trading simulation method can be implemented, for example, as software used in the information processing device 1. Moreover, the device for implementing the energy trading simulation method does not have to be a single device. In the case of multiple devices, they may be located at one location, or may be a system in which devices are located at remote locations and work together.

[0023] In the following, for example, a consumer model refers to a program created to reflect the characteristics of a consumer, or the execution of such a program, and includes an order behavior agent. Also, for example, an order behavior agent refers to a software module that is part of a consumer model, etc., and autonomously performs judgments, decisions, actions, etc. The same applies to other models and agents. In the following, for the sake of convenience of explanation, the subject of the action of an agent may also be described as a model.

[0024] As shown in FIG. 1, a consumer model of multiple sellers (such as consumer A model Ma), a consumer model of multiple buyers (such as consumer B model Mb), a P2P power auction model Mc, and a market maker model Mm are assumed here.

[0025] The consumer A model Ma includes a sell order behavior agent Aa. The consumer B model Mb includes a purchase order behavior agent Ab. The P2P power auction model Mc includes a trading agent Ac. The market maker model Mm includes a market making agent Am.

[0026] Here, the storage unit 2 and the processing unit 6 in Fig. 2 will be described. The storage unit 2 stores a plurality of business operator models (consumer A model Ma, consumer B model Mb) and a trading model (P2P power auction model Mc) for simulating buying and selling transactions of power products for each time period into which a day is divided. The storage unit 2 also stores various other information related to buying and selling transactions of power (auction trading conditions (trading fees, trading units), various parameters, etc.).

[0027] The processing unit 6 includes an order behavior agent processing unit 61 , a trading agent processing unit 62 , and a control unit 63 .

[0028] The order behavior agent processing unit 61 executes the processing of the order behavior agents (selling order behavior agent Aa, buying order behavior agent Ab) based on a business operator model (consumer model, etc.). For example, the order behavior agent processing unit 61 operates each of a plurality of order behavior agents that determine and output the buying / selling type, order price, and order quantity of electricity trading based on input of electricity supply and demand forecast (electricity demand forecast, electricity supply forecast), market trading price, and trading agreement performance, using each of a plurality of business operator models. Hereinafter, when describing the processing by the order behavior agent processing unit 61, the operating subject may be the ordering behavior agent.

[0029] The trading agent processing unit 62 uses a trading model (P2P power auction model Mc) to determine and output the market trading price and the trading agreement performance of each of the multiple ordering behavior agents based on the input of the buying / selling type, order price, and order quantity output by each of the multiple ordering behavior agents.

[0030] For example, the order behavior agent simulates changes in at least one of the types of electricity trading, the order price, and the order quantity when there is a change in the expected electricity supply and demand and / or a change in the forecast power generation volume of a power source that uses renewable energy.

[0031] For example, the ordering behavior agent further outputs an estimated revenue value of the energy trading. Here, the revenue refers to the difference between the revenue from the P2P energy auction and the revenue from a normal transaction (buying and selling with a retailer).

[0032] Also, for example, the order behavior agent updates the investment behavior characteristics according to the magnitude of the estimated revenue from electricity trading, based on the investment behavior characteristics of order behavior related to trend following and counter-trend trading in the operator model.

[0033] Also, for example, the trading agent Ac simulates multiple trading formats (opening (itayose) and auction (zaraba)). In this case, the order behavior agent updates the trading allocation characteristics according to the magnitude of the estimated revenue of the energy trading based on the trading allocation characteristics of the order behavior regarding the order ratio to the multiple trading formats in the business operator model.

[0034] Also, for example, the order behavior agent outputs a comparison result of revenue estimates associated with one or more differences among the power source mix, investment behavior characteristics, and transaction allocation characteristics.

[0035] For example, the trading agent Ac further outputs indicator information regarding market liquidity.

[0036] The trading agent also simulates trading of multiple electricity products that differ in one or more of the following: the ratio of electricity derived from renewable energy sources in the electricity, the CO2 emission intensity, and the region where the electricity is generated.

[0037] The trading agent also outputs the results of a comparison of indicator information regarding market liquidity associated with one or more differences among the presence or absence of a market maker, the trading unit, the trading fee, and the price limit.

[0038] A more specific example will be described below.

[0039] The consumer A model Ma is a model of a consumer in which the (predicted) renewable energy power generation output within consumer A is greater than the power demand for the target time of a specified auction trading product k (for example, product k=22, 10:30-11:00 kWh of electricity), and surplus electricity is expected to be generated. The consumer A model Ma inputs the amount of electricity (kWh) of the predicted renewable energy output, the estimated amount of electricity demand (kWh), and behavioral characteristic values ​​(market selection characteristic, order price characteristic, continuous order characteristic).

[0040] In addition, in the case of simulated conditions in which the weather forecast information changes during the auction time and the amount of power in the renewable energy output forecast or the estimated power demand changes, the values ​​after the change are input for the amount of power in the renewable energy output forecast and the estimated power demand. In addition, the consumer A model Ma inputs the auction price and the contract result of the consumer A model Ma's order (whether or not there is a contract, the contract price if there is a contract, and the contract amount) from the P2P power auction model Mc.

[0041] The consumer A model Ma also outputs the marginal cost characteristics, the order price of the sell order calculated by the sell order behavior agent Aa, and the order quantity (kWh) to the P2P power auction model Mc. Here, the marginal cost characteristics reflect the transaction fee of the P2P power auction. The order price and order quantity reflect the transaction price increment (tick size) and transaction unit (kWh) of the P2P power auction. The consumer A model Ma also calculates an estimate of the revenue (estimated revenue) in the P2P power auction transaction based on the order execution result of the consumer A model Ma and the marginal cost characteristics.

[0042] On the other hand, the consumer B model Mb is a model of a consumer (including consumers who do not have power generation facilities) whose renewable energy output (forecasted value) is smaller than the power demand during the same specified power trading time period and who wants to purchase renewable energy power at a low price to make up for the power shortage. The configuration of the consumer B model Mb is similar to that of the consumer A model Ma, but differs in that it outputs the marginal utility characteristics, the order price of the buy order, and the order quantity (kWh) calculated by the buy order behavior agent Ab to the P2P power auction model Mc.

[0043] For example, 100 or more seller consumer models and buyer consumer models can be used. Also, the number of seller consumer models and buyer consumer models does not need to be the same.

[0044] In addition, the P2P electricity auction model Mc determines the matching of buy and sell orders for each type of electricity trading, opening (Itayose) and auction (Zaraba), and outputs the trading price and the order contract results for each consumer model (whether or not there is a contract, the contract price if there is a contract, and the contract amount).The P2P electricity auction model Mc also calculates and outputs the volatility of the auction price (price volatility) and an index of auction liquidity (an example of an index related to market liquidity) based on the auction contract results.

[0045] Examples of indicators related to market liquidity include the following (1) to (3). (1) Contract volume for one trading period (kWh) (2) Average bid-ask spread (yen) = Average value of the gap between the best bid and best ask over one trading period (3) Price range / volume ratio (yen / kWh) = (difference between maximum and minimum contract prices in one trading period) ÷ (contract amount in one trading period)

[0046] Moreover, the market maker model Mm receives order book information for buy orders and sell orders from the P2P power auction model Mc, and outputs both buy orders and sell orders.

[0047] In addition, the control unit 63 executes processes other than those performed by the order behavior agent processing unit 61 and the trading agent processing unit 62.

[0048] 3 is a flowchart showing the process of simulating energy trading by the information processing device 1 of the first embodiment. In the first half, the process related to the opening (itayose) is performed, and in the second half, the process related to the auction (zaraba) is performed.

[0049] First, in step S1, the processing unit 6 inputs information on the auction transaction conditions (transaction fee, transaction unit) from the storage unit 2.

[0050] In steps S2 to S22, the processing unit 6 performs processing for each date (d=1, . . . , Nd).

[0051] In step S3, the processing unit 6 selects the behavioral characteristics (market selection characteristic, order price characteristic, continuous order characteristic) of each consumer model.

[0052] In steps S4 to S20, the processing unit 6 performs processing for each auction trading item (k=1, . . . , Nk).

[0053] In steps S5 to S9, the processing unit 6 performs processing for each consumer model (i=1, . . . , Ni).

[0054] In step S6, the order action agents (selling order action agent Aa, buying order action agent Ab) perform renewable energy output prediction and power demand estimation.

[0055] Next, in step S7, the ordering behavior agent selects marginal cost characteristics and marginal utility characteristics.

[0056] Here, Fig. 4 is an explanatory diagram of (a) the cost characteristic of the seller consumer model and (b) the utility characteristic of the buyer consumer model in the first embodiment. A method for selecting the cost characteristic of the consumer A model Ma and the utility characteristic of the consumer B model Mb will be described with reference to Fig. 4.

[0057] As shown in Figure 4(a), if consumer A model Ma, a seller of surplus electricity, is able to sell electricity to retailers at unit price Ps, there is no economic benefit to selling electricity through P2P trading at a price lower than Ps, including the commission for P2P trading. Therefore, the marginal cost characteristics are drawn on the higher side than the unit price Ps. Note that the amount of surplus electricity includes forecast errors, so as it approaches the surplus electricity forecast amount Q, the marginal cost curve rises to reflect the risk of being disappointed.

[0058] On the other hand, as shown in Figure 4(b), if a renewable energy power buyer, Consumer B Model Mb, can purchase electricity from a retailer at unit price Pb, there is no economic benefit to purchasing electricity through P2P trading at a price above Pb, including the commission for P2P trading. Therefore, the marginal utility curve is drawn on the side lower than the unit price Pb. Similarly, the amount of electricity demand also contains forecast errors, so as it approaches the expected amount of electricity demand, the marginal utility curve falls to match the risk of the forecast being inaccurate.

[0059] Returning to FIG. 3, after step S7, in step S8, the order behavior agent processing unit 61 places an opening order.

[0060] After steps S5 to S9, in step S10, trading agent processing unit 62 performs contract processing (item-based order processing). For example, it simulates an order-based order processing method in which the cost characteristics and utility characteristics taking into account the order price characteristic α (described later) of each consumer model are synthesized and the contract price is determined from the intersection point.

[0061] Next, in steps S11 to S19, the processing unit 6 performs processing for each auction trading time (t=0, . . . , Nt).

[0062] In steps S12 to S18, the processing unit 6 performs processing for each consumer model (i=1, . . . , Ni).

[0063] In step S13, the order behavior agent performs renewable energy output prediction and power demand estimation.

[0064] Next, in step S14, the ordering behavior agent calculates the base price.

[0065] In addition, the market selection characteristic β (0≦β≦1) of Consumer A model Ma represents the proportion of orders placed at the opening of the market among the order quantity that matches the surplus energy amount Q obtained by subtracting the expected demand from the renewable energy output. The order quantity for the surplus energy amount (1-β)Q that was not ordered at the opening of the market and the orders that were not executed at the opening of the market are placed in the continuous auction.

[0066] In addition, the order price characteristic (book-based price characteristic) α (0≦α≦1) indicates that in book-based transactions, in the case of sellers, the order price is increased by 1 / (1-α) times based on the marginal cost characteristic (Figure 4(a)), and in the case of buyers, it indicates the discount rate from the order price based on the marginal utility characteristic (Figure 4(b)).

[0067] In addition, the continuous order characteristic γ (0≦γ≦1) represents the ratio of the technical characteristics out of the fundamental characteristics, which place an order at a contrarian price level based on the reference price Pf, and the technical characteristics, which place an order at a time that follows the auction trading price trend. Additionally, behavioral characteristics such as minimum order unit quantity may be entered for each ordering behavior agent.

[0068] After step S14, in step S15, the ordering behavior agent places an order in the continuous auction. Here, the ordering behavior in the continuous auction by the ordering behavior agent will be described with reference to FIG.

[0069] Fig. 5 is an explanatory diagram of a method of calculating a reference price for continuous orders according to the contract volume performance in the first embodiment. In the case of the sell order behavior agent Aa, a target trajectory of the amount of electricity sold is set so that the transaction is completed within the auction time, and when the amount of electricity sold contracted is less than the target trajectory (points P1 to P2 in Fig. 5(a)), the contract is prioritized and the reference price is lowered. Also, when the amount of electricity sold contracted is more than the target trajectory (points P2 and after in Fig. 5(a)), a characteristic of prioritizing profitability and raising the reference price is simulated.

[0070] This is simulated by the feedback control in Figure 5(b). In other words, the reference price Pf(t) is calculated by PI control (P: Proportional, I: Integral) using the difference between the contracted amount of electricity sold and the target trajectory of the amount of electricity sold. Note that PID control (D: Differential) or P control may be used instead of PI control.

[0071] In addition, we simulate the ordering behavior of each customer based on the expected return according to the continuous auction price P(t). The expected return Re(t) and the order price Ro i (t) is as follows:

number

[0072] Note that the first term on the right hand side of the expected return Re(t) is the expected return for fundamental characteristics, and the second term is the expected return for technical characteristics.

[0073] In the case of the buying order behavior agent Ab, the order price characteristics are as shown in Fig. 6. Fig. 6 is a diagram showing the relationship between continuous order characteristics and order prices in the first embodiment. Fig. 7 is an explanatory diagram of a continuous auction model in the first embodiment. Fig. 8 is a diagram showing an example of a simulation result of a continuous auction in the first embodiment.

[0074] The buyer's customer model is assumed to place orders in order (in the order of agent numbers) at regular intervals.

[0075] 3, in step S16, the market making agent Am of the market maker model Mm places an order. For example, the market making agent Am outputs both a sell order and a buy order at an order price with a predetermined price spread based on the midpoint between the buy order and the sell order, based on the order book information of the buy order and the sell order in the P2P power auction model Mc.

[0076] Next, in step S17, the trading agent Ac of the P2P power auction model Mc performs continuous contract processing. For example, the trading agent Ac simulates a continuous auction method in which, based on the orders of each consumer model, individual orders are matched according to the principles of price priority and time priority, as shown in Fig. 7, and trades are contracted as needed. Fig. 7(a) shows a continuous auction model without a market maker model. Fig. 7(b) shows a continuous auction model with a market maker model.

[0077] Figure 8 shows an example of a simulated continuous auction. In the simulations ((c) and (d)) that assume a time when there is a large amount of surplus renewable energy power, the auction price is lower than in the simulations ((a) and (b)) that do not.

[0078] After steps S4 to S20 in FIG. 3, in step S21, each consumer model calculates an estimate of the profit in the P2P power auction transaction (estimated profit) based on the contract result of the order and the like.

[0079] 3, in step S23, the trading agent Ac calculates the volatility of the auction price and auction liquidity information (spread, depth, etc.) based on the simulation results of the contract price, contract amount, or order price during the auction trading period. This ends the explanation of FIG. 3.

[0080] The consumer model updates its behavioral characteristics so as to improve the estimated profit R, simulating consumer learning. As an example, the market selection characteristic β is updated using the gradient algorithm shown in the following formula. R(d) and R(d-1) are the estimated profits in the P2P auction for the same time on the current day and the previous day, respectively, and Δβ is the update width. Δβ may be different for each consumer model. Also, sign(·) is a sign function that returns a value of -1, 0, or 1 depending on the sign of the number in (·).

number

[0081] Furthermore, an order behavior agent that simulates algorithmic trading that automatically places orders based on preset rules may be constructed and introduced into the simulation. Furthermore, an algorithm that learns using AI (Artificial Intelligence) may be introduced into the order behavior agent.

[0082] It is also possible to simulate electricity trading by conducting auctions for renewable energy and non-renewable energy electricity products, auctions for electricity products generated in different regions, and auctions for electricity products with multiple trading units including small lots (for example, 0.1 kWh instead of 1 kWh).

[0083] As described above, according to the first embodiment, even when the actual data of the energy trade is insufficient, the energy trade can be simulated with high accuracy.

[0084] Even in cases where there is no trading history and no data, it is possible to simulate electricity trading with high accuracy by using multiple order behavior agents that combine the seller's cost characteristics and buyer's utility characteristics, which are constructed based on economic principles, with the behavioral characteristics of traders.In addition, since the order behavior agents calculate the base price for orders based on the contract volume history, it is possible to simulate auction price changes that reflect the supply and demand balance of the entire auction.

[0085] In addition, by updating the behavioral characteristics of the ordering behavior agent so as to improve the estimated profit based on the estimated profit results for each predetermined period, it is possible to simulate energy trading including traders who learn and change their ordering behavior based on trading experience. Furthermore, by comparing the relationship between the behavioral characteristics of the ordering behavior agent and the estimated profit and estimating behavioral characteristics for increasing profitability, it is possible to support each trader in energy trading.

[0086] It is also possible to simulate electricity trading, including simulating changes in trading prices in P2P electricity auctions that reflect changes in solar power output forecasts and changes in electricity demand forecasts during the trading time.

[0087] It is also possible to simulate and estimate the magnitude of relative changes in price volatility and auction liquidity when the minimum unit or transaction fee of P2P energy trading auction is changed or when a market maker is introduced. Furthermore, it can be useful in supporting the system design of P2P energy auction to ensure auction liquidity.

[0088] Furthermore, if the order behavior agent is configured to output (display) the comparison results of revenue estimates associated with differences in one or more of the power source composition, investment behavior characteristics, and transaction allocation characteristics, this can serve as a reference for knowing which parameters affect revenue and to what extent.

[0089] Second embodiment Next, a second embodiment will be described. For matters similar to those in the first embodiment, duplicated explanations will be omitted as appropriate. FIG. 9 is a schematic diagram showing an overview of the simulation of power trading in the second embodiment. FIG. 10 is an explanatory diagram of the cost characteristics of the power generation business operator model and the utility characteristics of the retail business operator model in the second embodiment.

[0090] In the second embodiment, as shown in Fig. 9, a power market transaction in which a plurality of power generation companies and retailers participate is simulated using a plurality of order behavior agents and Itayose and Zaraba execution calculations. Specifically, the models include a power generation company A model M1, a power generation company B model M2, a retailer A model M3, a retailer B model M4, and a power market model M5.

[0091] The power generation company A model M1 includes a sell order behavior agent Ad. The power generation company B model M2 includes a sell order behavior agent Ae. The retailer A model M3 includes a purchase order behavior agent Af. The electricity market model M5 includes a trading market agent Ag (an example of a trading agent).

[0092] Each power generation business model inputs renewable energy output forecasts, power source and cost characteristics, and behavioral characteristics (market selection characteristics, order price characteristics, continuous order characteristics). Each power generation business model then outputs the estimated revenue of each power generation business based on the simulation results. The retail business model is almost the same as the buyer consumer model in Figure 1, so a description is omitted.

[0093] The cost characteristics and utility characteristics of the power generation business model and the retail business model can be assumed as shown in Figure 10. As shown in Figure 10(a), the cost characteristics of the power generation business model can be defined as the unit price of electricity increasing in a step-like manner as the amount of generated electricity increases. As shown in Figure 10(b), the utility characteristics of the retail business model can be defined as the unit price of electricity decreasing in a linear manner as the estimated amount of electricity demand increases.

[0094] When performance data is available, the power cost characteristics and behavioral characteristics of each business operator may be determined so as to match the market price fluctuation characteristics.

[0095] In this way, according to the second embodiment, by using multiple ordering behavior agents that combine the cost characteristics of power producers and the utility characteristics of retailers, which are constructed based on economic principles, with the behavioral characteristics of traders, it is possible to realize an electricity trading simulation that reflects both the electricity supply and demand balance and the behavioral characteristics of trading companies. In addition, since each ordering behavior agent calculates the base price for orders based on the contract volume record, it is possible to simulate market price changes according to the supply and demand balance of the entire market.

[0096] It is also possible to realize a simulation of electricity trading, including a simulation of changes in market prices of electricity trading that reflect changes in predicted photovoltaic power output and changes in predicted electricity demand during the trading time.

[0097] It is also possible to simulate and estimate the relative impact of the power generation mix of power generation companies and the behavioral characteristics of retail companies on the profitability of each company. Furthermore, by simulating the power generation mix of power generation companies and the behavioral characteristics of retail companies (market selection, order price characteristics, continuous order characteristics) to increase profitability in electricity trading, it is possible to support each company in their electricity trading.

[0098] The program executed by the information processing device 1 of this embodiment may be provided by being recorded in an installable or executable format on a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, or a DVD (Digital Versatile Disk).

[0099] The program may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. The program may be provided or distributed via a network such as the Internet.

[0100] Moreover, the program may be provided in a state where it is pre-installed in a ROM or the like.

[0101] Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0102] 1...information processing device, 2...storage unit, 3...input unit, 4...display unit, 5...communication unit, 6...processing unit, 61...order behavior agent processing unit, 62...trading agent processing unit, 63...control unit, Aa, Ad, Ae...sell order behavior agents, Ab, Af...buy order behavior agents, Ac...trading agent, Ag...trading market agent, Ma...consumer A model, Mb...consumer B model, Mc...P2P electricity auction model, Mm...market maker model, M1...power generation business operator A model, M2...power generation business operator B model, M3...retailer A model..., M4...retailer B model, M5...electricity market model

Claims

1. a storage unit that stores a plurality of business operator models for simulating buying and selling transactions of electric power products for each time period obtained by dividing a day into a plurality of time periods, and the trading models; A plurality of order action agents are operated to determine and output a buying / selling type, an order price, and an order quantity of an electricity transaction by a Proportional Integral Differential (PID) control calculation, which is a calculation for bringing the transaction result closer to a transaction target corresponding to the electricity supply and demand forecast, based on the input of the electricity supply and demand forecast, the market transaction price, and the transaction contract result, corresponding to each of the plurality of business operator models; a processing unit that operates a trading agent that determines and outputs the market trading price and the trading contract performance of each of the plurality of order behavior agents based on the input of the buy / sell type, the order price, and the order quantity output by each of the plurality of order behavior agents in accordance with the trading model; Equipped with The processing unit uses the trading contract performance output by the trading agent as an input for the PID control calculation during the next operation of the order behavior agent.

2. the order behavior agent simulates a change in at least one of the buying and selling type, the order price, and the order quantity of the electricity transaction when there is a change in at least one of an expected change in electricity supply and demand and a change in a forecast of a power generation amount of a power source that uses renewable energy; 2. The energy trading simulator according to claim 1, wherein the trading agent simulates a trading market including continuous trading.

3. The energy trading simulator according to claim 1 or 2, wherein the order behavior agent further outputs a revenue estimate of the energy trading.

4. 4. The energy trading simulation device according to claim 3, wherein the order behavior agent updates the investment behavior characteristics according to the magnitude of the estimated profit of the energy trading based on the investment behavior characteristics of the order behavior related to trend following and counter-trend trading in the operator model.

5. The trading agent simulates a plurality of trading formats including a stock market transaction and a continuous market transaction, 5. The energy trading simulation device according to claim 4, wherein the order behavior agent updates a market selection characteristic, which is a trading allocation characteristic of order behavior related to order ratios to the plurality of trading types in the company model, according to the magnitude of the estimated revenue of the energy trading by using a gradient method algorithm or a predetermined learning algorithm using AI (Artificial Intelligence).

6. 6. The energy trading simulator according to claim 5, wherein the order behavior agent outputs a comparison result of the revenue estimate accompanying a difference when a change is specified for one or more of a power source composition, the investment behavior characteristic, and the market selection characteristic.

7. The energy trading simulator according to claim 1 , wherein the trading agent further outputs indicator information regarding market liquidity.

8. 8. The energy trading simulator according to claim 7, wherein the trading agent simulates a plurality of electricity commodity transactions that differ in one or more of the ratio of renewable energy-derived electricity in the electricity, the CO2 intensity, and the region where the electricity is generated.

9. 8. The energy trading simulator according to claim 7, wherein the trading agent outputs a comparison result of index information regarding the market liquidity associated with one or more differences among the presence or absence of a market maker, a trading unit, a trading fee, and a price limit.

10. A method for simulating buying and selling of electric power products for each time period obtained by dividing a day into a plurality of time periods, comprising: a first step of operating a plurality of order action agents, each of which corresponds to each of the plurality of business operator models, to determine and output a buying / selling type, an order price, and an order quantity of an electricity transaction by a PID control calculation, which is a calculation for bringing the transaction contract results closer to a transaction target corresponding to the electricity supply and demand forecast, based on an input of an electricity supply and demand forecast, a market transaction price, and a transaction contract result; a second step of operating a trading agent that determines and outputs the market trading price and the trading contract performance of each of the plurality of order behavior agents based on the input of the buy / sell type, the order price, and the order quantity output by each of the plurality of order behavior agents in accordance with the trading model; Including, Repeating the first step and the second step; the trading contract performance output by the trading agent in the second step is used as an input for the PID control calculation during operation of the order behavior agent in the next first step.

11. A computer having a storage unit that stores a plurality of business operator models for simulating buying and selling transactions of electric power products for each time period obtained by dividing a day into a plurality of time periods, Operate each of a plurality of order action agents that determine and output a buying / selling type, an order price, and an order quantity of an electricity transaction by a PID control calculation, which is a calculation for bringing the transaction contract results closer to a transaction target corresponding to the electricity supply and demand forecast, based on an input of an electricity supply and demand forecast, a market transaction price, and a transaction contract result, corresponding to each of a plurality of said business operator models; a program for causing a trading agent to function as a processing unit that determines and outputs the market trading price and the trading contract performance of each of the plurality of order behavior agents based on input of the buying / selling type, the order price, and the order quantity output by each of the plurality of order behavior agents in accordance with the trading model, The processing unit uses the transaction contract performance output by the trading agent as an input for the PID control calculation during the next operation of the order behavior agent.

Images