JP2021005404A - Information processor, method for processing information, and program - Google Patents

Abstract

To provide a bidding price determination device for determining an efficient trade reference price in a market before an electric power time.SOLUTION: A bidding price determination device 1 includes: a market information acquisition unit 100; a product information feature amount extraction unit 120; a market price prediction unit 130; a trade reference price determination unit 140; and an output unit 160. The market information acquisition unit acquires a price history of a product before a current time in the market before a an electric power time. The product information feature amount extraction unit extracts a first feature amount as the feature amount of the price history of the product from the acquired price history. The market price prediction unit predicts a second feature amount as the feature amount of a price transition of the product after the current time on the basis of the extracted first feature amount. The trade reference price determination unit determines a trade reference value as the reference of a trace of the product after the curent time on the basis of the predicted second feature amount.SELECTED DRAWING: Figure 1

Description

Embodiments of the present invention relate to information processing devices, information processing methods and programs.

The pre-hour market, which opened in April 2016, is the final market for selling and procuring electricity, and electricity can be bought and sold from 17:00 the day before until one hour before the actual delivery of electricity. .. The pre-hour market is held in a rough manner. Considering that the electric power spot market, which is the conventional method, is a single price auction, an algorithm different from the electric power trading in the spot market is required to make an automatic bid in the pre-electric power market. Unlike financial products such as stocks, it is difficult to store electric power to be traded in a large capacity, and its absolute amount also fluctuates depending on the date and time. For this reason, when dealing only with electric power products, it is difficult to form a portfolio like financial products, and the need for efficient automatic trading is increasing.

In order to solve this problem, a technology for predicting the price transition of the market before the hour by simulating the spot market and the real-time market on the previous day and a conventional technology for creating a draft of the electricity transaction price have been devised. However, the technology for predicting price transitions is large-scale simulation and may take too much time. In addition, the technology for creating a draft electricity transaction price is for bilateral transactions and cannot be used as it is in the pre-hour market.

Japanese Patent No. 4087735 Japanese Unexamined Patent Publication No. 2006-9432

Provides an information processing device that determines an efficient reference price for trading in the pre-power market.

According to one embodiment, the bid price determination device includes an acquisition unit, an extraction unit, a prediction unit, a determination unit, and an output unit. The acquisition unit acquires the price history of the goods in the pre-power market market before the current time. The extraction unit extracts the first feature amount, which is the feature amount of the price history of the product, from the acquired price history. The prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time of the product, based on the extracted first feature amount. The determination unit determines a transaction reference value that serves as a reference for transactions after the current time of the product, based on the predicted second feature amount. The output unit outputs the determined transaction reference value.

The block diagram which shows the function of the bid price determination device which concerns on one Embodiment. The figure which shows the product and the trading period in the market before the electric power time. A flowchart showing the processing of the bid price determination device according to the embodiment. The figure which shows the relationship between the price history and the bid which concerns on one Embodiment. The graph which shows the example of the price history which concerns on one Embodiment. A graph showing an example of bidding according to one embodiment. A block diagram showing the functionality of another example of a bid pricing device. A flowchart illustrating the processing of another example of a bid pricing device. A block diagram showing the function of yet another example of a bid pricing device.

Hereinafter, embodiments will be described in detail with reference to the drawings.

The bid price determining device according to the present embodiment acquires the trading history of the product in the market before the power hours of the previous day, predicts the characteristic quantity of the product on the current day from the characteristic quantity of the trading history, and obtains the predicted feature quantity. Based on this, the bid price of the product on the day is determined.

FIG. 1 is a diagram showing an example of a block diagram showing a function of the bid price determination device 1 according to the present embodiment. As shown in FIG. 1, the bid price determination device 1 includes a market information acquisition unit 100, a market information storage unit 110, a product information feature amount extraction unit 120, a market price prediction unit 130, and a transaction standard price determination unit 140. A market information monitoring unit 150 and an output unit 160 are provided. The bid price determination device 1 has a function of acquiring various data from the electric power exchange 2 and automatically bidding on the electric power exchange 2, or a function of presenting a reference price to be automatically bid to the user.

The bid price determination device 1 determines the transaction price of each product for a plurality of products sold and sold on the electric power exchange 2. The goods bought and sold on the electric power exchange 2 will be described. FIG. 2 is a diagram showing commodities traded on the electric power exchange 2 in one day.

The product refers to the electric power delivered every 30 minutes by dividing the 24 hours from 0:00 to 24:00 every 30 minutes. That is, as shown in FIG. 2, for example, the electric power delivered from 0:00 to 0:30 is the product 1, and the electric power delivered from 0:30 to 1:00 is the product 2, ..., 23:00. The electric power delivered from 30 minutes to 24:00 (0:00 the next day) can be the product 48.

These commodities are traded (sold and sold) in a rough-field format from 17:00 on the previous day to one hour before the delivery of the commodities. That is, for example, the product 1 can be traded between 17:00 on the previous day and 23:00 on the previous day, and the product 2 can be traded between 17:00 on the previous day and 23:30 on the previous day. Similarly, the commodity 48 can be traded between 17:00 on the previous day and 22:30 on the current day. In this way, each product has a separate trading period, and each product is traded within a predetermined period.

Hereinafter, in the description in the present specification, words relating to dates such as the previous day, the current day, and the next day are based on the consumption (supply and demand) time of electric power as a product (current day). That is, for the product 1, the product 2, and the product 3, the electric power trading period on that day exists on the previous day. For example, the trading period of the product 1 on the day is from 17:00 on the previous day to 23:00 on the previous day. On the other hand, for products 4 and later, the electric power trading period on the day extends from the previous day to the current day. For example, the trading period of the product 48 is from 17:00 on the previous day to 22:30 on the current day. Further, the timing of processing is based on the timing of bidding, for example, the current time.

Returning to FIG. 1, the market information acquisition unit 100 acquires the transaction price history of each product in the market before the current time published by the electric power exchange 2, for example, via a network. For example, when the current time is 18:00 on the current day, for product 1, the transaction history from 17:00 on the previous day to 23:00 on the previous day is acquired as price history information, and from 17:00 on the current day to 18:00 on the current day. Get the transaction history as the latest price information. For product 48, the transaction history from 17:00 the day before to 22:30 the day before is acquired as price history information, and the transaction history from 17:00 the day before to 18:00 on the day is acquired as the latest price information. ..

The information acquired by the market information acquisition unit 100 is stored in the market information storage unit 110. The market information storage unit 110 includes a price history information storage unit 112 and a price latest information storage unit 114. The price history information storage unit 112 stores the price history information among the information acquired by the market information acquisition unit 100. The price latest information storage unit 114 stores the latest price information among the information acquired by the market information acquisition unit 100. In this way, for a certain product to be traded on the day, the latest price history information that has already been closed is stored in the price history information storage unit 112, and the latest history information of the price that has not been closed is the latest price information. It is stored in the storage unit 114. The market information storage unit 110 is provided in the bid price determination device 1 in FIG. 1, but is not limited to this, and may be provided in another server or the like via a network, for example.

The product information feature amount extraction unit 120 extracts the first feature amount necessary for predicting the future bid price from the history information stored in the price history information storage unit 112. The first feature amount is a feature amount indicating a feature in the history information of the transaction of the product. The extracted first feature amount is output to the market price prediction unit 130. As another example, the market information acquisition unit 100 may calculate the first feature amount in advance from the acquired history information and store it in the price history information storage unit 112. In this case, the product information feature amount extraction unit 120 extracts the first feature amount for the required period from the price history information storage unit 112.

The market price prediction unit 130 is required to determine the transaction reference price (transaction reference value) that is the basis for predicting the bid price based on the first feature amount extracted by the product information feature amount extraction unit 120. 2 Features are predicted using, for example, a prediction model. The second feature amount is a feature amount that indicates the characteristics of future price fluctuations of the product. The predicted second feature amount is output to the transaction reference price determination unit 140.

The forecast model used is, for example, a model generated by performing regression analysis with the first feature as the explanatory variable and the second feature as the objective variable, using the past price history of the product in advance. is there. This regression model may be any model as long as it can appropriately predict the second feature amount from the first feature amount. For example, the regression model may be a linear model such as a linear regression or a regression with penalties (for example, ridge regression, Lasso), a non-linear model such as a random forest or a neural network, or other than the examples. It may be another model. As the estimation method, for example, an estimation method such as a least squares method or a Bayesian estimation can be appropriately selected according to the above model. As yet another example, a Kalman filter or the like may be used for dynamic fitting.

As the past price history of the product used when generating the prediction model, for example, the price history information for the past one month is used. For the price history information for the past one month, a model between the first feature amount of the previous day and the second feature amount of the current day is generated. The past one month is just an example, and as another example, a model may be generated for the same month, the same season, etc. one year ago. As yet another example, data from the past few years may be used to generate a model with heavier weights closer to the trading day. In this way, it is possible to use a model that can appropriately predict the second feature amount on the trading day.

The transaction reference price determination unit 140 determines the transaction reference price (transaction reference value) which is the reference value of the bid price based on the second feature amount predicted by the market price prediction unit 130. The transaction reference price is a reference price for comparison with the price of the product at each timing when actually bidding. The determined transaction reference price is output to the market information monitoring unit 150. The conversion from the second feature quantity to the transaction reference price may be a linear conversion or a non-linear conversion, and is preset by, for example, a simulation using past market information. As another example, the conversion standard may be set so as to be adjusted so that the value is defeated from the second feature amount to the safe side.

The market information monitoring unit 150 monitors the latest price information stored in the price latest information storage unit 114, and compares the latest price of the product with the transaction standard price determined by the transaction standard price determination unit 140. Then, it is determined whether or not each timing is suitable for bidding. This determination result is output to the output unit 160.

The output unit 160 outputs to the outside based on the determination result of the market information monitoring unit 150. For example, the output unit 160 includes a bidding unit (not shown) and a display unit. The bidding department determines whether or not to bid at each timing based on the judgment result of the market information monitoring unit 150, and trades with the electric power exchange 2 at a timing suitable for bidding, for example, via a network. Bid automatically at a price based on the base price. The display unit notifies the user of the transaction standard price, etc. by displaying the transaction standard price, etc. Instead of this display unit, a printing unit for printing, a communication unit for outputting necessary information by communication, an audio output unit for outputting voice, and the like may be provided.

Next, the flow of processing of the above-described configuration, for example, the flow up to bidding (buying order) for the product 2 on the side of procuring electric power will be described more specifically with reference to FIGS. 3 and 4. FIG. 3 is a flowchart showing a processing flow of the bid price determining device 1 according to the present embodiment. FIG. 4 is a diagram showing the time flow of the transaction of the product 2. Since it is a sale and purchase of the product 2, the timing of all processing is assumed to be the time between 17:00 and 23:30 on the previous day in the following description. The unit of price is basically [yen / kWh], but for the sake of simplicity, it is simply described as [yen].

First, the market information acquisition unit 100 acquires the price history information of the product published by the electric power exchange 2 (S100). Specifically, as shown in FIG. 4, the transaction price history (price history information) and the previous day from 17:00 on the day before the previous day to 23:30 on the day before the previous day, which is the transaction period on the day before the product 2. The price history (latest price information) of the market on the day from 17:00 to the timing when the market information acquisition unit 100 acquires the information is acquired.

The price history information need only be acquired once after the market for the product is closed the day before. For example, the transaction history of the previous day that can be acquired by 17:00 of the previous day at the timing of 17:00 of the previous day, that is, all the transaction histories from product 1 to product 33 (or product 34) of the previous day are acquired and the price history information is stored. It may be stored in the unit 112. The remaining products may be acquired as soon as the price history becomes available. Alternatively, each product may be acquired when the market is closed. Further, the information stored in the price latest information storage unit 114 may be converted into history information. It is preferable to appropriately acquire the latest price history at each timing.

Next, the product information feature amount extraction unit 120 extracts the first feature amount from the price history of the product 2 stored in the price history information storage unit 112 (S102). For example, the opening price, the high price, the low price, and the closing price of the product 2 are extracted as the first feature quantity from the price history of the market on the previous day. FIG. 5 is a diagram that briefly shows the price transition of a product in the market on a certain day. For example, in the case of the price transition as shown in FIG. 5, the opening price is 8.8 yen, the high price is 9.0 yen, the low price is 8.0 yen, and the closing price is 8.4 yen as the first feature amount.

Next, the market price prediction unit 130 predicts the second feature amount based on the extracted first feature amount (S104). The second feature quantity is, for example, a feature quantity such as a high price or a low price in the market on the day. The prediction of the second feature amount is calculated using a prediction model using the first feature amount as an explanatory variable. In the following, for example, it is assumed that the low price is set as the second feature amount and the market price forecasting unit 130 predicts that the second feature amount is 7.5 yen.

Next, the transaction reference price determination unit 140 determines the transaction reference price based on the predicted second feature amount (S106). The transaction reference price determination unit 140 determines the transaction reference price, for example, by linearly converting the predicted second feature quantity. As an example of the linear conversion, the transaction reference price = a × the second feature amount + b may be used. More specifically, assuming that a = 1.1 and b = 0.0, the transaction base price is determined as 7.5 × 1.1 + 0.0 = 8.25 yen.

After the transaction reference price is determined, the transaction reference price may be displayed and output to the user via a display or the like provided in the output unit 160 (S108). At least one of the first feature amount and the second feature amount may be displayed together with the display of the transaction reference price. Further, when the first feature amount is output, it may be output together with which past data the first feature amount is based on.

By outputting in this way, the user can recognize what kind of feature amount in the historical information at what time in the past was used to determine the transaction reference price. It is also possible to confirm whether the transaction standard price and the feature amount used to determine the transaction standard price are appropriate. If the user thinks that this result is not valid, the user may manually set parameters and the like. Here, the user may be, for example, a side that purchases electric power or a side that supplies electric power.

When displaying as described above, the product information feature amount extraction unit 120, the market price prediction unit 130, and the transaction reference price determination unit 140 have extracted the first feature amount, the predicted second feature amount, and the determined transaction reference price, respectively. May be directly notified to the output unit 160. By doing so, the user can acquire the transaction reference price of the product desired by the user and the information of the first feature amount and the second feature amount used for the determination thereof in real time.

Further, not only the second feature quantity such as the low price but also other state quantities of the market on the day may be displayed. For example, not only the information on the low price in the market on the day but also the predicted value of the high price and the predicted value of the closing price may be output. In this way, it is also possible to output other state quantities required for forecasting transactions.

Further, a prediction model of the second feature amount used, a correction formula (correction model) of the transaction reference price, and the like may be displayed. By outputting in this way, the user can recognize based on what model the transaction base price is determined.

On the other hand, the output unit 160 may bid on the electric power exchange 2 using the determined transaction reference value. In this case, an automatic bid is made by loop processing based on the determined transaction reference price and the current bid price of the product (S110). This loop processing is repeated, for example, until the timing when the required amount of electric power can be secured or the timing when the market closes before the time.

When bidding, first, the market information monitoring unit 150 acquires the latest transaction price of the product stored in the price latest information storage unit 114 (S112). In this case, the information stored in the price latest information storage unit 114 is not necessarily read out, and may be acquired from the electric power exchange 2 via the market information acquisition unit 100 at a necessary timing. In such a case, the price latest information storage unit 114 is not an essential configuration.

Next, the market information monitoring unit 150 compares the determined transaction reference price with the latest transaction price of the product (S114), and determines whether or not the product satisfies the transaction condition at the timing (S114). S116). The transaction condition is satisfied, for example, when the user intends to procure the product 2, the latest transaction price is lower than the transaction reference price.

In addition, the present invention is not limited to this, and the transaction reference price may be adjusted based on the current timing of the market before the time and the timing until the closing. For example, when the closing time is short, adjustment may be made so as to judge that the price is suitable for trading (purchase) even if it is slightly higher than the trading standard price. As another example, the trading base price may be dynamically changed based on the time remaining until closing.

Adjustments to these transaction base prices or adjustments to the transaction base price itself should be made so that the transaction base price from opening to closing is adjusted using vertically bounded functions such as the sigmoid function and the inverse trigonometric function. Alternatively, adjustments may be made so that the transaction base price increases linearly from the opening to the closing. Not limited to these, linear and non-linear adjustments can also be made. Further, the price may be adjusted based not only on the timing of opening and closing but also on the amount of electric power required to be procured at the timing and the amount of electric power already procured at the timing.

When the market price at that timing satisfies the transaction condition (S116: YES), the output unit 160 bids on the electric power exchange 2. For example, if the low price in the latest price information of the acquired product 2 is lower than the transaction standard price of 8.25 yen, a buy order is placed. The order may be placed using the trading market price as the limit price or may be placed at the market price.

As a result of bidding, if the procured electric power has not yet reached the required amount and if the transaction conditions are not satisfied (S116: NO), the processes of S112 to S118 are repeated. On the other hand, when the required amount of electricity can be procured, the process exits the loop of S110 to S118 and ends the process.

The bid may be made by repeatedly looping a predetermined amount of electricity until the amount of electricity procured reaches the required amount, or the amount of electricity required in one bid may be bid. Alternatively, at the timing when the transaction base price is determined, the bid may be made with the transaction base price as the limit price without loop processing. The market information monitoring unit 150 is not an indispensable configuration when bidding at the limit price at the timing when the transaction standard price is determined. In this way, the timing of bidding and the amount of electricity are not limited to these as long as the necessary amount of electricity can be procured appropriately.

FIG. 6 is a diagram showing an example of price transition of the market before the electric power time of the product. When the market on the day changes as shown in FIG. 6, the output unit 160 automatically bids at the timing when the market price falls below the transaction standard price of 8.25 yen (the period indicated by the diagonal line). As described above, the timing of bidding and the amount of bids may be determined. For example, when bidding for a predetermined amount many times, the bid should be made once and then after a predetermined period of time. You may. In this way, for example, in the case of a price transition as shown in FIG. 6, a bid is made at a price lower than 8.25 yen, but there is a high possibility that a bid can be made at a lower price.

As described above, according to the present embodiment, in the pre-power market, by bidding at the reference price predicted based on the past (for example, the previous day) history information of the product, the pre-power market When procuring electricity, it becomes possible to procure electricity at a price close to the expected low price. In other words, it is possible to automatically trade at an appropriate price in the pre-power market where it is difficult to generate a portfolio.

(First modification)
In the above, the procurement of electric power has been described, but it is also possible to bid for the sale of electric power. When selling electric power, for example, a regression model is used to predict a high price as the second feature amount, and the transaction reference price is determined using a linear model such as 0.9 x second feature amount + 0.0 [yen]. To.

In the situation described above, when the user is the seller, the second feature amount is predicted to be, for example, 9.5 yen, and in this case, the transaction base price is 0.9 × 9.5 + 0.0 = 8. It becomes .55 [yen]. Then, the bid price determination device 1 can automatically make a bid to sell electricity at an amount of 8.55 yen or more or 8.55 yen or more by using this transaction reference price. .. Of course, the price is not limited to 8.55 yen, and the transaction base price may be dynamically changed by dynamically estimating the second feature amount or correcting the transaction base price as described above.

As described above, it is possible to use the bid price determination device 1 according to the present embodiment not only when procuring electric power but also when selling electric power in the market before electric power hours.

(Second modification)
The price history information and the latest price information are not limited to the above. For example, even if the price history information is not closed, the information that is not the latest information is stored from the price latest information storage unit 114. It may be transferred to unit 112 and stored. By doing so, it becomes possible to reflect the price history that has not yet closed in the market on the day in the determination of the transaction base price. In this case, the first feature quantity may be, for example, extracted from the price history of the market on the day before closing and the price history of the market on the day when it is not closed, such as high price and low price. When the market price history of the day is used in this way, the transaction reference price may be updated at the timing when the first feature amount is changed.

In particular, as the product number increases, the trading period in the market before the hour of the day becomes longer, and in addition to the transactions in the market before the time of the previous day, the importance of the transaction information already performed in the market of the day becomes higher. come. In such a case, the accuracy of the bid price determination at the procurement timing can be further improved by reflecting the transaction history in the market before the closing time of the day in the forecast of the second feature quantity and the determination of the transaction base price. It is possible to improve. When incorporating the market information of the day into the forecast and decision, the forecast model of the second feature quantity and the correction model of the transaction reference price may be different from the case of using the information up to the previous day, or as described above. Similarly, it may be calculated by a dynamic model.

(Third modification example)
The first feature is not limited to the four prices of opening price, high price, low price and closing price, but other statistics such as mean value, variance value, standard deviation value, median value and mode value in the price transition of the previous day market. It may be a quantity. In this case, it is possible to predict the second feature amount by setting the explanatory variables in the regression model corresponding to the extracted first feature amount. For example, the second feature amount may be predicted by using three values of the mean value, the variance value, and the median value as explanatory variables. As another example, in addition to the four values, the average value and the variance value of the price may be taken into consideration, and these six values may be used as explanatory variables to obtain the second feature amount.

(Fourth modification)
The bid price determination device 1 may include an environmental information acquisition unit (not shown) to acquire the temperature and weather of the day. A model may be generated as an explanatory variable for predicting the second feature amount by combining the acquired temperature and weather with the above-mentioned first feature amount. Further, these environmental information may be used not only for the prediction model of the second feature amount but also for the correction when obtaining the transaction reference price from the second feature amount. In this case, in addition to the past market information, a correction model including information such as past temperature and weather may be generated.

(Fifth modification)
The model for predicting the second feature amount may also use the first feature amount of the product before and after the product. For example, when predicting the second feature amount of the product 2, in addition to the first feature amount of the product 2, the first feature amount of the product 1 and the product 3 may be input as an explanatory variable. In this case, the model may be such that the weight of the first feature amount of the product 2 is heavier than the weight of the first feature amount of the product 1 and the product 3. Further, not only the product 1 and the product 3, but also the product 4 and the latest product 48, etc., may be expanded in the range of extracting the input first feature amount. In this case, the weighting for each product when predicting the second feature amount may be changed.

(6th modification)
In the above-described embodiment and modification, the transaction reference price determination unit 140 has described an example in which the transaction reference price is determined by making a linear or non-linear correction to the second feature amount, but with a more nonparametric distribution. It may be corrected as if it exists.

The case of procuring the product 2 will be described. For example, assume that the second feature amount is an average value (average contract price) and is predicted to be 8.5 yen. The transaction standard price determination unit 140 generates a model in advance by assuming the average value as the median value by simulation and analysis using past market information, and assuming that the value of the 20th percentile is the transaction standard price from this average value. You may. Alternatively, the second feature amount may be predicted as the median value, and the value of the 20th percentile may be used as the transaction reference price. Specifically, for example, the distribution of the transaction price of the product 2 is acquired from the market information up to the previous day. Based on this distribution information, the price relationship of the 20th percentile is calculated from the median price of the past market, and the transaction base price is determined based on this relationship. This relationship may be expressed using a function. This function can be linear or non-linear. As an example, it is possible to function as a linear model expressing the median value and the value of the 20th percentile in the distribution information up to the previous day as a ratio.

For example, if 8.5 yen is substituted in the obtained function and 8.25 yen is output, the transaction base price is set to 8.25 yen. By using this transaction reference price, it is possible to carry out efficient automatic bidding as in the above-described embodiment and each modification. By setting the average value to the 80th percentile, the transaction reference price can be determined from the high price, and this transaction reference price can be used when selling electric power.

(7th modification)
The bid price determination device 1 described above does not particularly accept input by the user, and determines the transaction reference price using the processing method and parameters set in the device in advance, but the present invention is not limited to this. .. That is, the user may be able to correct the transaction reference value by referring to the transaction reference price or the like output from the output unit 160.

FIG. 7 is a diagram showing a bid price determination device 1 according to this modification. The bid price determination device 1 according to the present modification further includes a price adjustment unit 170 in the bid price determination device 1 according to the above-described embodiment and each modification. Further, FIG. 8 is a flowchart showing the flow of processing according to this modification.

When the user wants to manually adjust the transaction reference price output by the output unit 160, the user adjusts the transaction reference price in the price adjustment unit 170 via an input interface (not shown). The price adjustment unit 170 receives an input from the user and adjusts the transaction reference price (S109). After that, the price adjustment unit 170 notifies the output unit 160 of the adjusted transaction reference price. Upon receiving the notification, the output unit 160 automatically makes a subsequent bid using the adjusted transaction reference price.

As described above, the user can adjust the output transaction reference value by further providing the price adjustment unit 170. By doing so, it is possible to efficiently perform automatic bidding and to set a price that the user is satisfied with. Further, even when the user wants to set the transaction standard price from the beginning, the bid price determination device 1 can present the standard value based on the past price history of the product to be traded, which is troublesome for the user. Can be omitted.

The value adjusted by the user is not limited to the transaction standard price. For example, when the output unit 160 outputs the second feature amount, the user may be able to adjust the output second feature amount. Then, the price adjustment unit 170 determines the transaction reference price using the second feature amount adjusted by the user, and the output unit 160 bids based on the adjusted transaction reference price determined by the price adjustment unit 170. You may do it.

(8th modification)
In the seventh modification described above, the price adjustment unit 170 for adjusting the transaction standard price, etc. by the user with reference to the output transaction standard value, etc. is provided, but the price adjustment method is limited to this. Absent.

FIG. 9 is a block diagram showing a function of the bid price determination device 1 according to this modification. The bid price determination device 1 further includes a processing method instruction unit 180. The processing method instruction unit 180 enables, for example, the product information feature amount extraction unit 120 to instruct the user of an extraction method for extracting the first feature amount. For example, the product information feature amount extraction unit 120 extracts the opening price, the high price, the low price, and the closing price of the market on the previous day as the first feature amount, as in the above-described embodiment.

When the user wants to change the feature amount to be extracted as the first feature amount, the processing method instruction unit 180 is provided with four feature amounts, for example, an average value, a dispersion value, a median value, and a mode value, as the first feature amount. Enter to extract. In response to this input, the processing method instruction unit 180 extracts the product information feature amount extraction unit 120 into four feature amounts of the mean value, the variance value, the median value, and the mode value as the first feature amount. Give instructions. After receiving the instruction, the product information feature amount extraction unit 120 shifts to the process of extracting the four new feature amounts as the first feature amount. Then, after that, the second feature amount is predicted based on the first feature amount, and then the transaction reference price is determined.

As described above, according to the present modification, the user can instruct the processing method at each stage, and it is possible to determine an appropriate bid price that more reflects the intention of the user.

Note that the processing method instruction unit 180 gives instructions not only to the product information feature amount extraction unit 120. That is, in the processing method instruction unit 180, the market price prediction unit 130 predicts the second feature amount from the first feature amount (prediction model, etc.), and the transaction reference price determination unit 140 sets the transaction reference price from the second feature amount. The determination method (correction model, etc.) for determining the above may be instructed.

(9th modification)
In the above-described embodiment and modification, the bid price determination device 1 that determines the transaction reference price based on the history of the transaction price in the pre-power market is described, but not only the price in the pre-power market as an explanatory variable. , Electric power spot market transaction information can also be used as an explanatory variable.

That is, the market information acquisition unit 100 acquires not only the transaction history of the market before the electric power time but also the transaction history of the electric power spot market. For example, the transaction price in the spot market on the previous day, which is the transaction period of the product 2 on the current day, is acquired. Specifically, information such as the contract price, the contract amount, the sell bid amount, and the buy bid amount of the spot market of the product 2 on the previous day is acquired.

The product information feature amount extraction unit 120 first obtains information on the spot market on the previous day, such as contract price and contract quantity, together with information on the market before the power hours on the previous day, for example, information on the opening price, high price, low price, and closing price. Extract as a feature quantity. At this time, the spot market information may be used as it is as the first feature amount, or may be processed and used as the first feature amount. Subsequent processing is the same as that of the above-described embodiment or modification. That is, the extracted first feature amount is used as an explanatory variable, the second feature amount which is the objective variable is predicted, and the transaction reference price is determined from the predicted second feature amount.

As described above, according to this modification, it is possible to predict the high price and the low price in the time-ahead market of the day with higher accuracy by acquiring the information of the spot market as well as the time-ahead market. .. In this way, by increasing the accuracy, it becomes possible to perform automatic bidding more efficiently than the above-described embodiment and each modification.

In addition, as in the above-mentioned modification, in order to further improve the accuracy, the spot market information of the products (product 1, product 3, etc.) before and after the product 2 or the spot market information before the previous day is acquired. You may.

In all the above-mentioned examples, what is shown as a magic number is only an example, and the embodiments and modifications are not bound to this. For example, the coefficients 1.1 and the values of the 20th percentile used in the explanation in the correction model are not limited to these.

At least a part of the analysis system and the analysis apparatus described in all the above-described embodiments may be configured by hardware or software. When composed of software, a program that realizes at least a part of the functions of the analysis system and the analysis device is stored in a storage medium such as a flexible disk or a CD-ROM, read by a computer, and executed. May be good. The storage medium is not limited to a removable one such as a magnetic disk or an optical disk, and may be a fixed storage medium such as a hard disk device or a memory.

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 embodiments, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are also included in the scope of the invention described in the claims and the equivalent scope thereof. Further, as a matter of course, it is also possible to partially and appropriately combine these embodiments within the scope of the gist of the present invention.

(Additional note)
The embodiments in the present disclosure can be expressed as the following embodiments.

(1)
An extraction unit that extracts the first feature quantity, which is the feature quantity, from the price history of the product before the current time in the market before the electric power time.
Based on the extracted first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time of the product, and
A determination unit that determines a transaction reference value that serves as a reference for transactions after the current time of the product based on the predicted second feature amount.
With
The extraction unit extracts the opening price, the high price, the low price, and the closing price in the price history of the product as the first feature amount in a predetermined period.
Information processing device.

(2)
The information processing apparatus according to (1), further comprising an acquisition unit for acquiring the price history.

(3)
The information according to (1) or (2), wherein the prediction unit predicts the high price or the low price of the product as the second feature amount based on the first feature amount of the product extracted by the extraction unit. Processing equipment.

(4)
The prediction unit predicts the average contract price of the product as the second feature amount based on the first feature amount of the product extracted by the extraction unit, according to any one of (1) to (3). The information processing device described.

(5)
The information processing apparatus according to any one of (1) to (4), wherein the determination unit linearly transforms the second feature amount predicted by the prediction unit to determine the transaction reference value.

(6)
The information processing apparatus according to any one of (1) to (4), wherein the determination unit determines the transaction reference value by non-linearly converting the second feature amount predicted by the prediction unit.

(7)
Any of (1) to (6) further including a display unit that displays at least one of the second feature amount predicted by the prediction unit and the transaction reference value determined by the determination unit as numerical data. The information processing device described in.

(8)
Further provided with an adjustment unit that accepts input from the user and adjusts the second feature amount or the transaction reference value.
After the display unit displays at least one of the second feature amount and the transaction reference value as numerical data to the user, the adjustment unit adjusts the second feature amount or the transaction reference value by input from the user. Then, when the second feature amount is adjusted, the transaction reference value is determined using the second feature amount, and when the transaction reference value is adjusted, the adjusted transaction reference value is used as described above. Use as the transaction standard value,
The information processing device according to (7).

(9)
The information processing apparatus according to any one of (1) to (8), further comprising a bidding section for bidding for the product in the pre-power market based on the transaction reference value.

(10)
It also has a monitoring unit that monitors the price of the goods in the pre-power market.
The bidding unit compares the transaction reference value with the price monitored by the monitoring unit, and bids based on the comparison result.
The information processing device according to (9).

(11)
The information processing device according to (10), wherein the bidding unit automatically bids based on the comparison result.

(12)
The information processing according to any one of (1) to (11), further comprising an instruction unit that receives a processing method used until the transaction reference value is determined from the user and instructs the processing method to be processed. apparatus.

(13)
The instruction unit accepts the method for extracting the first feature amount as the processing method, and extracts the first feature amount from the price history using the extraction method for the first feature amount received by the extraction unit. Instructed to
The extraction unit extracts the first feature amount based on the extraction method.
The information processing device according to (12).

(14)
The instruction unit receives the method for predicting the second feature amount as the processing method, and uses the prediction method for the second feature amount received by the prediction unit to obtain the second feature amount from the first feature amount. Instruct to predict,
The prediction unit predicts the second feature amount from the first feature amount based on the prediction method.
The information processing device according to (12) or (13).

(15)
The instruction unit accepts the method for determining the transaction reference value as the processing method, and determines the transaction reference value from the second feature amount using the determination method for the transaction reference value received by the determination unit. Instruct to
The determination unit determines the transaction reference value from the second feature amount based on the determination method.
The information processing apparatus according to any one of (12) to (14).

(16)
A process in which the extraction unit extracts a first feature amount, which is a feature amount of the price history of the product, from the price history of the product before the current time of the product in the pre-power market.
Based on the extracted first feature amount, the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time of the product.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time of the product based on the predicted second feature amount.
With
The first feature amount is an opening price, a high price, a low price, and a closing price in a predetermined period in the price history of the product.
Information processing method.

(17)
Computer,
A means for extracting a first feature quantity, which is a feature quantity of the price history of the product, from a price history of the product before the current time of the product in the market before the electric power time.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time of the product, based on the extracted first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time of the product based on the predicted second feature amount.
To function as
The first feature amount is an opening price, a high price, a low price, and a closing price in a predetermined period in the price history of the product.
program.

(18)
An extraction unit that extracts the first feature quantity, which is the feature quantity, from the price history of the product before the current time in the market before the electric power time.
Based on the extracted first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time of the product, and
A determination unit that determines a transaction reference value that serves as a reference for transactions after the current time of the product based on the predicted second feature amount.
A display unit that displays at least one of the second feature amount predicted by the prediction unit and the transaction reference value determined by the determination unit as numerical data.
An adjustment unit that receives input from the user, displays at least one of the second feature amount or the transaction reference value as numerical data, and then adjusts the second feature amount or the transaction reference value.
Information processing device equipped with.

1: Bid price determination device, 100: Market information acquisition unit, 110: Market information storage unit, 120: Product information feature quantity extraction unit, 130: Market price prediction unit, 140: Transaction standard price determination unit, 150: Market information monitoring Unit, 160: Output unit

Claims (33)

An extraction unit that extracts the first feature quantity, which is the feature quantity of the price history before the current time of the product in the market before the electric power time,
Based on the first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time,
Based on the second feature amount, a determination unit that determines a transaction reference value that serves as a reference for transactions after the current time, and
With
The extraction unit extracts at least one of the opening price, the high price, the low price, and the closing price of the price history in a predetermined period as the first feature amount.
Information processing device. An extraction unit that extracts the first feature quantity, which is the feature quantity of the price history before the current time of the product in the market before the electric power time,
Based on the first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time,
Based on the second feature amount, a determination unit that determines a transaction reference value that serves as a reference for transactions after the current time, and
With
The prediction unit predicts a high price or a low price as the second feature amount based on the first feature amount extracted by the extraction unit.
Information processing device. An extraction unit that extracts the first feature quantity, which is the feature quantity of the price history before the current time of the product in the market before the electric power time,
Based on the first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time,
Based on the second feature amount, a determination unit that determines a transaction reference value that serves as a reference for transactions after the current time, and
With
The prediction unit predicts the average contract price as the second feature amount based on the first feature amount extracted by the extraction unit.
Information processing device. An extraction unit that extracts the first feature quantity, which is the feature quantity of the price history before the current time of the product in the market before the electric power time,
Based on the first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time,
Based on the second feature amount, a determination unit that determines a transaction reference value that serves as a reference for transactions after the current time, and
With
The determination unit linearly transforms the second feature amount to determine the transaction reference value.
Information processing device. An extraction unit that extracts the first feature quantity, which is the feature quantity of the price history before the current time of the product in the market before the electric power time,
Based on the first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time,
Based on the second feature amount, a determination unit that determines a transaction reference value that serves as a reference for transactions after the current time, and
With
The determination unit determines the transaction reference value by non-linearly converting the second feature amount.
Information processing device. The extraction unit extracts at least one of the opening price, the high price, the low price, and the closing price of the price history in a predetermined period as the first feature amount.
The information processing device according to any one of claims 2 to 5. The prediction unit predicts a high price or a low price as a second feature amount based on the first feature amount extracted by the extraction unit.
The information processing apparatus according to claim 1, or any of claims 3 to 6. The prediction unit predicts the average contract price as the second feature amount based on the first feature amount extracted by the extraction unit.
The information processing apparatus according to claim 1, claim 2, or any of claims 4 to 7. The determination unit linearly transforms the second feature amount to determine the transaction reference value.
The information processing device according to any one of claims 1 to 3 or 5 to 8. The determination unit determines the transaction reference value by non-linearly converting the second feature amount.
The information processing device according to any one of claims 1 to 4, or 6 to 9. Acquisition department that acquires the price history,
The information processing apparatus according to any one of claims 1 to 10, further comprising. A display unit that displays at least one of the second feature amount and the transaction reference value as numerical data, and
The information processing apparatus according to any one of claims 1 to 11, further comprising. An adjustment unit that accepts input from the user and adjusts the second feature amount or the transaction reference value.
With
After the display unit displays at least one of the second feature amount and the transaction reference value as data to the user, the adjustment unit adjusts the second feature amount or the transaction reference value by input from the user. When the second feature amount is adjusted, the transaction reference value is determined using the second feature amount, and when the transaction reference value is adjusted, the adjusted transaction reference value is used for the transaction. Use as a reference value
The information processing device according to claim 12. The information processing apparatus according to any one of claims 1 to 13, further comprising a bidding section for bidding in the pre-power market market based on the transaction reference value. It also has a monitoring unit that monitors the price of the goods in the pre-power market.
The bidding unit compares the transaction reference value with the price monitored by the monitoring unit, and bids based on the comparison result.
The information processing device according to claim 14. The information processing device according to claim 15, wherein the bidding unit automatically bids based on the comparison result. The information processing according to any one of claims 1 to 16, further comprising an instruction unit that receives a processing method used until the transaction reference value is determined from the user and instructs the processing method to be processed. apparatus. The instruction unit accepts the method for extracting the first feature amount as the processing method, and extracts the first feature amount from the price history using the extraction method for the first feature amount received by the extraction unit. Instructed to
The extraction unit extracts the first feature amount based on the extraction method.
The information processing device according to claim 17. The instruction unit receives the method for predicting the second feature amount as the processing method, and uses the prediction method for the second feature amount received by the prediction unit to obtain the second feature amount from the first feature amount. Instruct to predict,
The prediction unit predicts the second feature amount from the first feature amount based on the prediction method.
The information processing apparatus according to claim 17 or 18. The instruction unit accepts the method for determining the transaction reference value as the processing method, and determines the transaction reference value from the second feature amount using the determination method for the transaction reference value received by the determination unit. Instruct to
The determination unit determines the transaction reference value from the second feature amount based on the determination method.
The information processing apparatus according to any one of claims 17 to 19. An extraction unit that extracts the first feature quantity, which is the feature quantity of the price history before the current time of the product in the market before the electric power time,
Based on the first feature amount, a prediction unit that predicts the second feature amount, which is the feature amount of the price transition after the current time,
Based on the second feature amount, a determination unit that determines a transaction reference value that serves as a reference for transactions after the current time, and
A display unit that displays at least one of the second feature amount and the transaction reference value as numerical data, and
An adjustment unit that receives input from the user, displays at least one of the second feature amount or the transaction reference value as numerical data, and then adjusts the second feature amount or the transaction reference value.
Information processing device equipped with. The processing that the extraction unit extracts the first feature amount, which is the feature amount of the price history before the current time of the product in the market before the electric power time,
A process in which the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time, based on the first feature amount.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
With
The first feature amount is at least one of an opening price, a high price, a low price, and a closing price in a predetermined period for the price history.
Information processing method. Computer,
A means for extracting the first feature quantity, which is the feature quantity of the price history before the current time of the product in the pre-power market.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time, based on the first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
To function as
The first feature amount is at least one of an opening price, a high price, a low price, and a closing price in a predetermined period for the price history.
program. The processing that the extraction unit extracts the first feature amount, which is the feature amount of the price history before the current time of the product in the market before the electric power time,
A process in which the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time, based on the first feature amount.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
With
Based on the first feature amount, a high price or a low price is predicted as the second feature amount.
Information processing method. Computer,
A means for extracting the first feature quantity, which is the feature quantity of the price history before the current time of the product in the pre-power market.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time, based on the first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
To function as
Based on the first feature amount, a high price or a low price is predicted as the second feature amount.
program. The processing that the extraction unit extracts the first feature amount, which is the feature amount of the price history before the current time of the product in the market before the electric power time,
A process in which the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time, based on the first feature amount.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
With
Based on the first feature amount, the average contract price is predicted as the second feature amount.
Information processing method. Computer,
A means for extracting the first feature quantity, which is the feature quantity of the price history before the current time of the product in the pre-power market.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time, based on the first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
To function as
Based on the first feature amount, the average contract price is predicted as the second feature amount.
program. The processing that the extraction unit extracts the first feature amount, which is the feature amount of the price history before the current time of the product in the market before the electric power time,
A process in which the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time, based on the first feature amount.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
With
The transaction reference value is determined by linearly converting the second feature amount.
Information processing method. Computer,
A means for extracting the first feature quantity, which is the feature quantity of the price history before the current time of the product in the pre-power market.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time, based on the first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
To function as
The transaction reference value is determined by linearly converting the second feature amount.
program. The processing that the extraction unit extracts the first feature amount, which is the feature amount of the price history before the current time of the product in the market before the electric power time,
A process in which the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time, based on the first feature amount.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
With
The second feature quantity is non-linearly converted to determine the transaction reference value.
Information processing method. Computer,
A means for extracting the first feature quantity, which is the feature quantity of the price history before the current time of the product in the pre-power market.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time, based on the first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
To function as
The second feature quantity is non-linearly converted to determine the transaction reference value.
program. The processing that the extraction unit extracts the first feature amount, which is the feature amount of the price history before the current time of the product in the market before the electric power time,
A process in which the prediction unit predicts the second feature amount, which is the feature amount of the price transition after the current time, based on the first feature amount.
A process in which the determination unit determines a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
A process in which the display unit displays at least one of the second feature amount and the transaction reference value as numerical data.
A process in which the adjusting unit receives an input from the user, displays at least one of the second feature amount or the transaction reference value as numerical data, and then adjusts the second feature amount or the transaction reference value.
Information processing method. Computer,
A means for extracting the first feature quantity, which is the feature quantity of the price history before the current time of the product in the pre-power market.
A means for predicting a second feature amount, which is a feature amount of a price transition after the current time, based on the first feature amount.
A means for determining a transaction reference value that serves as a reference for transactions after the current time based on the second feature amount.
A means for displaying at least one of the second feature amount and the transaction reference value as numerical data.
A means for adjusting the second feature amount or the transaction reference value after receiving input from the user and displaying at least one of the second feature amount or the transaction reference value as numerical data.
A program that functions as.

Images