JP6965864B2 - Electricity business profit and loss analysis system - Google Patents

Description

The present invention is an invention relating to an electric power business profit / loss analysis system for calculating future profit / loss of an electric power company.

The business of generating and selling electric power consists of purchasing most of the fuel from overseas, converting these fuels into electric power with a generator, and selling the electric power according to contracts and fluctuations in demand.

Estimates of future business profits and losses for electric power companies operating such businesses include uncertainties such as fuel prices, exchange rates that convert yen into currency for fuel purchase, and climate change that is strongly correlated with electricity demand. It is greatly affected by fluctuations in factors. Therefore, profit / loss risk quantified in consideration of these uncertain factors is necessary as an index for estimating future business profit / loss to be used at the time of business planning of such an electric power company. To quantify profit and loss risk, a large number of scenarios are generated from the model by modeling uncertain factors and Monte Carlo simulation, profit and loss are calculated for all scenarios, and then a large number of profit and loss calculation results are aggregated and analyzed. There is a way.

Patent Document 1 describes a method of calculating profit / loss risk of an electric power company with different parameter sets (= portfolios) and comparing them. For example, the profit and loss risk is calculated for portfolio A and portfolio B, and the superiority or inferiority of the portfolio is judged by comparing the calculation results.

Japanese Patent Application Laid-Open No. 2008-59125

When considering multiple types of uncertain factors (for example, fuel price and exchange rate), a scenario is a scenario of changes in the value predicted in the future for one uncertain factor, and a case is a collection of scenarios for each type of uncertain factor (for example). For example, one case consists of one exchange scenario and one fuel price scenario), and the profit and loss calculation in each case is performed by implementing the power generation plan in the same way for all the cases generated by the prior technology. Is implemented to quantify profit and loss risk.

Therefore, when it is desired to consider various uncertain factors, when the number of cases is increased for the purpose of improving calculation accuracy, when quantifying profit / loss risk in a plurality of portfolios as in Patent Document 1, calculation is performed. Since the time-consuming power generation plan is implemented for the number of cases, there is a problem that the calculation time required for quantifying the profit / loss risk increases.

From the data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric power business, the scenario generation unit that generates N scenarios for future forecasts related to uncertain factors, and the assumption of power supply and demand based on the scenarios. The power generation planning department that calculates M profits and losses for each scenario, the model creation department that uses the relationship between the M scenarios and profits and losses as a profit and loss model, and the power generation planning department that calculates the profit and loss model (NM). The risk is calculated from the abbreviated calculation unit that substitutes the individual scenarios and calculates the profit and loss for each scenario, the M profit and loss calculated by the power generation planning unit, and the (NM) profit and loss calculated by the abbreviated calculation unit. It is an electric business profit and loss analysis system characterized by having a risk amount calculation unit and a result display unit that displays the profit and loss risk calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

In addition, a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric power business, a scenario generation unit that generates N scenarios for future forecasts related to uncertain factors, and a demand and supply of electricity based on the scenarios. A power generation planning unit that calculates M profits and losses for each scenario from assumptions, a division model creation department that divides the relationship between the M scenarios and profits and losses into a profit and loss model that divides the demand for electricity by the capacity of the generator, and a profit and loss model. The abbreviated calculation unit that calculates the profit and loss for each scenario by substituting the (NM) scenarios that were not calculated by the power generation planning department, and the M profit and loss and the abbreviated calculation unit that were calculated by the power generation planning department. An electric business profit and loss analysis system characterized by having a risk amount calculation unit that calculates risk from (NM) pieces of profit and loss, and a result display unit that displays the risk of profit and loss calculated by the risk amount calculation unit. Is. However, N and M are natural numbers, and there is a relationship of M <N.

Since the abbreviated calculation unit is included as a component, the calculation time required for calculating the profit and loss in all cases and quantifying the profit and loss risk can be shortened.

It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 1. FIG. It is a block diagram of the electric power business profit and loss analysis system in Embodiment 1. It is a flowchart which shows the operation of the electric power business profit and loss analysis system in Embodiment 1. FIG. It is a figure for demonstrating the distinction between a scenario and a case in Embodiment 1. FIG. It is a figure for demonstrating the daily profit / loss for each case in Embodiment 1. FIG. It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 2. It is a flowchart which shows the operation of the electric power business profit and loss analysis system in Embodiment 2. It is a figure for demonstrating the division model in Embodiment 2. FIG. It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 3. It is a flowchart which shows the operation of the electric power business profit and loss analysis system in Embodiment 3. It is a figure for demonstrating the order of the scenario in Embodiment 3. FIG. It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 4. It is a flowchart which shows the operation of the electric power business profit and loss analysis system in Embodiment 4. It is a figure for demonstrating the recalculation target in Embodiment 5. It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 5. It is a flowchart which shows the operation of the electric power business profit and loss analysis system in Embodiment 5. It is a figure for demonstrating the convergence test in Embodiment 6. It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 6. It is a flowchart which shows the operation of the electric power business profit and loss analysis system in Embodiment 6. It is a block diagram which shows the function of the electric power business profit and loss analysis system in Embodiment 7.

Embodiment 1.
FIG. 1 is a block diagram of an electric utility profit / loss analysis system according to the first embodiment of the present invention. The electric power business profit / loss analysis system uses the data set in the data setting 10 to generate a large number of scenarios related to uncertain factors in the scenario generation unit 12 in order to quantify the future profit / loss risk of the electric power business, and generates a power generation plan. Part 13 calculates the power generation plan, profit and loss, etc. in each case. At this time, if the power generation planning unit 13 calculates the profit and loss for all cases, it takes a long time to process the power generation planning department 13, so it cannot be tolerated in actual operation depending on the number of uncertain factors and the calculation period of the profit and loss risk. Become.

Therefore, in the first embodiment, the profit and loss in each case is obtained at a higher speed than that of the power generation planning unit 13 by the functions of the model creation unit 14 and the omission calculation unit 15. When the profit / loss calculation for all cases is completed, the risk amount calculation unit 16 aggregates the profit / loss calculation results, and the result display unit 17 displays the results. The data storage unit 11 receives data from the data setting unit 10 and inputs / outputs data between the scenario generation unit 12, the power generation planning unit 13, the model creation unit 14, the abbreviated calculation unit 15, and the risk amount calculation unit 16. Then, the calculation result is output to the result display unit 17.

FIG. 2 shows the configuration of a general hardware device. The computer 1 is composed of a CPU 2, a main storage device 3, an auxiliary storage device 4, and an external storage device 5. The external storage device 5 may be connected via the network 6. The output device 7 displays the output result from the computer 1, and the input device 8 inputs information to the computer 1.

In the first embodiment of the present invention, the data setting unit 10 shown in FIG. 1 is, for example, any one of the input device 8, the main storage device 3, the auxiliary storage device 4, the network 6, and the external storage device 5 shown in FIG. Alternatively, it is realized by combining a plurality. For example, the data setting unit 10 inputs or acquires data related to profit / loss risk calculation, and inputs or acquires data to a storage device (for example, at least one of a main storage device 3, an auxiliary storage device 4, and an external storage device 5) which is a data storage unit 11. save. The data related to the profit / loss risk calculation includes, for example, the actual data referred to in the calculation in the scenario generation unit 12, the reference period of the actual data, the number of cases to be generated, the calculation period of the profit / loss risk, and the fuel cost of the generator in the power generation planning unit 13. Functions, operation constraints, the number of cases used for model generation in the model creation unit 14, and the like.

The data storage unit 11 is realized by at least one of the main storage device 3, the auxiliary storage device 4, the network 6, and the external storage device 5 shown in FIG. 2, for example, the data setting unit 10, the scenario generation unit 12, and the power generation planning unit. 13. Stores the calculation results of the model creation unit 14, the abbreviated calculation unit 15, and the risk amount calculation unit 16.

The scenario generation unit 12 can be realized by using the CPU 2 as the internal arithmetic processing of the computer 1. The scenario generation unit 12 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3 which is the data storage unit 11, processes the scenario generation unit 12, and sets the scenario related to the uncertain factor obtained as a result to the main storage device 3, the auxiliary storage device 4, and the external device. Store in at least one of the storage devices 5. Further, the scenario may be displayed on an output device 7 such as a display and shown to the operator. Scenarios related to uncertainties are, for example, exchange price movements with a length of profit / loss risk calculation period, electricity demand, and the like.

The power generation planning unit 13 can be realized by using the CPU 2 as the internal arithmetic processing of the computer 1. The power generation planning unit 13 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the power generation planning unit 13, and transfers the power generation plan result obtained as a result to at least one of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. save. Further, the power generation plan result may be displayed on an output device 7 such as a display and shown to the operator. The power generation plan result is, for example, a generator start / stop plan having a length of the profit / loss risk calculation period, a generator output, and the like.

The model creation unit 14 can be realized by using the CPU 2 as the internal arithmetic processing of the computer 1. The model creation unit 14 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the model creation unit 14, and stores the model obtained as a result in at least one of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. .. Further, the model may be displayed on an output device 7 such as a display and shown to the operator. The model is, for example, a linear regression model in which an uncertain factor is an explanatory variable and profit or loss is an explained variable.

The abbreviated calculation unit 15 can be realized by using the CPU 2 as the internal arithmetic processing of the computer 1. The abbreviated calculation unit 15 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, performs the processing of the omitted calculation unit 15, and calculates the profit / loss calculation result for the scenario obtained as a result of at least the main storage device 3, the auxiliary storage device 4, and the external storage device 5. Save to either. Further, the calculation result of the profit and loss for the scenario may be displayed on the output device 7 such as a display and shown to the operator.

The risk amount calculation unit 16 can be realized by using the CPU 2 as the internal calculation processing of the computer 1. The risk amount calculation unit 16 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the risk amount calculation unit 16, and transfers the risk amount obtained as a result to at least one of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. save. Further, the amount of risk may be displayed on an output device 7 such as a display and shown to the operator. The amount of risk includes, for example, the average and variance of profit and loss, VaR representing the maximum loss at a certain confidence level, and the like.

The result display unit 17 is at least one of the power generation planning unit 13, the model creation unit 14, the omitted calculation unit 15, and the risk amount calculation unit 16 stored in at least one of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. Display the result of either calculation.

FIG. 3 is a flowchart showing the operation of the electric power business profit / loss analysis system according to the first embodiment. Hereinafter, the operation will be described with reference to FIG.

First, the data setting unit 10 sets data related to profit / loss risk calculation (step S100). The data related to the profit / loss risk calculation are, for example, the actual data (exchange, fuel price, demand, etc.) referred to in the calculation in the scenario generation unit 12, and the reference period R ([day], [month], [year] of the actual data. Etc.), the number of cases to be generated N [pieces], the calculation period T of profit / loss risk ([day], [month], [year], etc.), the fuel cost function of the generator in the power generation planning unit 13, and the operation constraint (minimum stop). Time, minimum start-up time, etc.), retail unit price [yen / kWh], and the number of cases M [pieces] used for model generation in the model creation unit 14. These parameters related to profit / loss risk calculation are stored in the data storage unit 11.

Next, the scenario generation unit 12 estimates model parameters related to uncertain factors (step S101). There are various types of mathematical models for generating scenarios. For example, if a GBM (Geometric Brownian Motion) model is used as a model for expressing an exchange rate, it can be expressed as in Eq. (1).

Here, μ: average logarithmic rate of change, σ: volatility, dz: standard normal random number (mean 0, variance 1), and P: exchange, x is a variable transformation of x = lnP.

In the estimation of the model parameters in step S101, the actual data set by the data setting unit 10 is referred to only the actual data reference period R, and the values are obtained by performing statistical analysis on the average logarithmic change rate μ and the volatility σ. .. Although the GBM model is illustrated here as a model of exchange, the model for generating a scenario regarding uncertain factors in the present invention does not have to be this model. Further, the model parameters may be obtained by statistical analysis or may be set by user input.

In step S102, the scenario generation unit 12 continues to generate a scenario and a case based on the model related to the uncertain factor obtained in step S101. Here, the scenario is a change in the value predicted in the future due to one uncertain factor when considering a plurality of types of uncertain factors (for example, fuel price and exchange rate). A case is a collection of scenarios for each type of uncertain factor. For example, one case will consist of one exchange scenario and one fuel price scenario.

For example, when a scenario related to exchange is generated using the GBM model of equation (1), the _{work of obtaining x 1} is performed over the calculation period T of profit and loss risk _{by generating the initial value x 0} and the standard normal random number dz. implement. Here, by repeating the number of generated cases N, N scenarios having a length of T with respect to the exchange can be obtained.

FIG. 4 shows an exchange and fuel price scenario generated with T = 365 and N = 5, and an example of arranging these as cases. The upper left side of Fig. 4 is the exchange scenario (vertical axis is the exchange rate [yen / dollar], the horizontal axis is the time axis at predetermined intervals, here day), and the lower left side is the fuel price scenario (vertical axis is the fuel price [dollar / dollar / dollar / dollar). Barrel], the horizontal axis is the time axis according to the predetermined interval, here 5 days), and the graph on the right side (the vertical axis is the profit and loss risk, the horizontal axis is the time axis according to the predetermined interval, here the day) ), One case is extracted and five cases are arranged (case 1 is shown in the upper row and case 5 is shown in the lower row, and the others are omitted). The time axis according to a predetermined interval may be plotted in units of 30 minutes or in units of one day, for example.

The scenario generation unit 12 will generate a value having a length equal to or longer than the profit and loss analysis period with respect to the uncertain factor. Cases related to these uncertain factors are stored in the data storage unit 11.

In steps S103 and S104, the start / stop plan of the generator and the output of the start / generator are determined for each case. Depending on the assumed number of generators and the length of the calculation period T of the profit / loss risk, the calculation time of steps S103 and S104 may be a practical problem. Therefore, in the first embodiment, this problem is avoided by performing steps S103 and S104 only M times, which is less than the number of generation cases N times.

In step S103, the power generation planning unit 13 determines the start / stop plan of the generator in the profit / loss risk calculation period T for each case. The start / stop plan is determined while observing the generator operation restrictions (minimum stop time, minimum start time, output change speed constraint, output upper / lower limit constraint, etc.), fuel consumption constraint, and the like.

Originally, the restrictions on generator equipment are the same in all cases. However, when demand and fuel prices are considered as uncertain factors, the total number of generators to be started and the type of generator change, so it is necessary to determine the start / stop plan and generator output for each case. .. As a determination method, for example, the use of dynamic programming, which is one of the mathematical programming methods, can be considered. These start / stop plans are stored in the data storage unit 11.

In step S104, the power generation planning unit 13 subsequently determines the output of the activated generator based on the start / stop plan determined in step S103 for each case. This output determination is determined so that the evaluation function is the minimum or maximum, and the evaluation function includes, for example, the power generation cost. Further, as an output determination method, for example, the use of the isoλ method can be considered.

After determining the generator output, the power generation cost over the period T is calculated from the output and the fuel cost function. In addition, the retail income over the period T is calculated from the demand and the retail unit price. From these, the profit and loss from the electric power business is defined as the difference between retail income and power generation cost and calculated. Data related to the generator output and profit / loss are stored in the data storage unit 11.

In the first embodiment, the start / stop plan (S103) and the output determination (S104) are performed separately, but they may be determined at the same time or may be determined by repeated calculation. Moreover, although the dynamic programming method and the equal λ method are illustrated, other methods may be used. The power generation planning unit 13 will calculate the profit and loss in each scenario from the assumption of demand and supply using the scenario.

In step S105, when steps S103 and S104 are performed M times, M profit / loss results having the same length T are obtained for each case. FIG. 5 is a diagram for explaining daily profit / loss for each case. FIG. 5 illustrates daily profit and loss in five cases. The vertical axis is the daily profit / loss, and the horizontal axis is the time axis at predetermined intervals, which is the day here.

In step S106, the model creation unit 14 creates a profit / loss model from M cases having a length of the profit / loss risk calculation period T and the profit / loss calculation result. The model creation unit 14 models the relationship between the scenario and the profit / loss. The profit / loss model is a model of the relationship between profit and loss with respect to uncertain factors. For example, when foreign exchange, fuel, and demand are assumed as uncertain factors, the foreign exchange at time t x _{1, t} , the fuel price x _{2, t} , It can be expressed by the linear regression model equation (2) with the demand x _{3, t} as the explanatory variable and the profit / loss y _{t as the explained variable.}

Here, a ₀ to a ₃ are model parameters. By calculating these model parameters using the M cases of the period T calculated before step S106 and the profit and loss calculation result of the period T corresponding to these cases, the relationship between the uncertain factor and the profit and loss can be obtained. Can be modeled.

In step S107, the abbreviated calculation unit 15 substitutes the remaining case (NM) scenarios into the profit / loss model created in step S106, thereby omitting the power generation planning unit 13 and calculating the profit / loss in each case. ..

The abbreviation calculation unit 15 obtains the profit and loss in the scenario by substituting the scenario into the created model.

Further, these profits and losses are stored in the data storage unit 11. When calculating profit and loss for all N generated cases, profit and loss calculation has already been completed for M in steps S103 and S104, so in this step, the remaining (NM) cases are omitted. To calculate profit and loss.

For example, when the total number of N is 1000, M is 50 and (NM) is 950. In order for the electric power company to manage the profit and loss risk, it is not enough to manage the profit and loss risk unless the number of N is sufficiently large. This is because when managing profit and loss risk, it is necessary to consider scenarios with large fluctuations due to the occurrence of an earthquake or the like. Further, if there is no sufficient difference between the M number and the (NM) number, the meaning of providing the omitted calculation unit 15 is halved.

Against this background, for example, there are the following methods for determining the number of M pieces for obtaining the profit and loss calculation results in all N cases within a desired calculation time. It takes A [seconds] per case for the power generation plan of steps S103 and S104, B [seconds] for modeling step S106, and C [seconds] per case for the abbreviated calculation of step S107, and the desired calculation time is R [ In the case of [seconds], M for obtaining the profit and loss calculation results in all N cases within the desired calculation time can be indicated by the sum of the calculation times of each step as shown in the equation (3).
AM + B + C (N−M) ≤ R (3)

Here, for example, when N: 1000 [times], A: 10 [seconds], B: 20 [seconds], C: 0.5 [seconds], R: 1000 [seconds], M ≦ 50.5. Become. In this case, M is 50 times and (NM) is 950 times in order to obtain a highly accurate calculation result within a desired time. If further speeding up is required, the number of Ms may be reduced from 50 times. As described above, the number and ratio of M and N cannot be unconditionally determined, but depend on the volume of calculation amount, the convenience of the electric power company, and the like.

In step S108, when step S107 is executed (NM) times, (NM) profit and loss calculation results having a length of period T are obtained.

In step S109, in the risk amount calculation unit 16, N profit / loss calculation results are obtained from the M profit / loss calculation results obtained in step S105 and the (NM) profit / loss calculation results obtained in step S107. Has already been obtained. In step S109, a desired risk index is calculated from the N profit and loss calculation results. Risk indicators include, for example, the average value and variance of profit and loss calculation results, VaR representing the maximum loss at a certain confidence level, and the like. These risk indicators are stored in the data storage unit 11.

In step S110, the result display unit 17 displays the calculation result stored in the data storage unit 11.

In the electric power business profit and loss analysis system according to the first embodiment, the profit and loss calculation in each case can be obtained at a higher speed than that of the power generation planning unit 13 by the functions of the model creation unit 14 and the omitted calculation unit 15, so that the entire system can be obtained. The calculation time in the above can be shortened as compared with the conventional case.

Further, since the power generation planning unit 13 uses the mathematical programming method, the optimum operating state of the generator is required in each case, but it takes time to solve the problem. When calculating the profit and loss of N cases in total and quantifying the profit and loss risk as in the first embodiment, the calculation of the power generation planning unit 13 is performed only M times, and the calculation results of these M times are used. The model creation unit 14 models the uncertain factors and the profit and loss only once, and the remaining NM times are omitted. Since the calculation unit 15 performs the profit and loss calculation only by substituting the numerical value of the scenario into the model formula. Compared with the case where the power generation planning unit 13 calculates the profit and loss for N cases, it is possible to obtain the profit and loss in all cases at high speed.

Further, since the omitted calculation unit 15 is included as a component, it is possible to skip the power generation plan that takes a long calculation time and calculate the profit / loss for each case. Therefore, the profit / loss in all cases is calculated and the calculation time required for quantifying the profit / loss risk is calculated. Can be shortened.

For example, in the Monte Carlo simulation, the time required for the Monte Carlo simulation can be shortened by modeling the relationship between uncertain factors and profit and loss, reducing the number of power generation plans, and calculating the profit and loss in all cases. By including the abbreviated calculation unit 15 as a component, the power generation plan that takes a long calculation time can be skipped, the profit / loss for the scenario can be calculated, and the calculation time required for quantifying the profit / loss risk can be shortened.

In the first embodiment, the electric power demand has been described as an example of the demand that can be supplied with energy by the generator, but other energy may be used, and thermal energy or the like may be considered.

As described above, the electric business profit and loss analysis system includes a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, and a scenario generation unit that generates N future forecast scenarios related to uncertain factors. , A power generation planning department that calculates M profits and losses for each scenario from the assumption of power supply and demand based on the scenario, a model creation department that uses the relationship between the M scenarios and profits and losses as a profit and loss model, and power generation in the profit and loss model. The abbreviated calculation unit that calculates the profit and loss for each scenario by substituting the (NM) scenarios that were not calculated by the planning unit, and the M profit and loss and the abbreviated calculation unit that were calculated by the power generation planning unit calculated (N). -M) It has a risk amount calculation unit that calculates the risk from the profit and loss, and a result display unit that displays the risk of the profit and loss calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

Embodiment 2.
In the first embodiment, the operation of the electric power business profit / loss analysis system for quickly obtaining the electric power business profit / loss in all cases generated for uncertain factors is shown. In order to obtain profit and loss at high speed, the relationship between uncertain factors and profit and loss was modeled. In the second embodiment, a method for improving the accuracy of this modeling will be described.

As shown in FIG. 6, the second embodiment includes a division model creation unit 20 instead of the model creation unit 14 of the first embodiment. In the drawings, those having the same reference numerals are the same or equivalent thereof, and this is common to the entire text of the specification and all the drawings of the drawings. Furthermore, the forms of the components appearing in the entire specification are merely examples and are not limited to these descriptions.

The division model creation unit 20 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the division model creation unit 20, and stores the model obtained as a result in at least one of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. do. Further, the model may be displayed on an output device 7 such as a display and shown to the operator. The model is, for example, a linear regression model in which an uncertain factor is an explanatory variable and profit / loss is an explained variable. In addition, the uncertain factors include the demand to be satisfied by power generation and heat generation, and the range of data used for modeling is divided according to the capacity of the generator to be started, and modeling is carried out for each division.

FIG. 7 is a flowchart showing the operation of the electric power business profit / loss analysis system according to the second embodiment. Hereinafter, the operation will be described with reference to FIG. 7. In FIG. 7, the same processing as in the first embodiment is assigned the same number, and the description thereof will be omitted.

In step S200 of FIG. 7, the division model creation unit 20 creates a profit / loss model from M cases having a length of the profit / loss risk calculation period T and the profit / loss calculation result. The profit / loss model is a model of the relationship between profit and loss with respect to uncertain factors. For example, when foreign exchange, fuel, and demand are assumed as uncertain factors, the foreign exchange at time t x _{1, t} , the fuel price x _{2, t} , It can be expressed by a linear regression model formula with demand x _{3, t} as the explanatory variable and profit / loss y _{t as the explained variable.}

Here, if the uncertain factors include the demand to be satisfied by power generation, the starting generator will change in consideration of the property that the average power generation unit price changes depending on the number and type of starting generators required to meet the demand. By dividing the range of data used for modeling according to the capacity of the model and modeling for each category, the accuracy of modeling regarding the relationship between demand and profit / loss is improved.

The division model will be described with reference to FIG. The vertical axis is profit / loss [10,000 yen], and the horizontal axis is demand [MWh]. FIG. 8 is a diagram plotting the relationship between M cases including scenarios related to electric power demand and profit / loss calculation results, focusing on electric power demand and profit / loss. The horizontal axis shows demand and the vertical axis shows profit and loss, and model formulas are obtained for each of the four categories from category A to category D. Here, the model formula has linearity, but the modeling (model formula) may be a non-linear model. In addition, one plot shows the profit and loss in the electric power demand in a certain case at t time.

Focusing on the profit and loss when the power demand is around 450 [MWh], if the range where the demand is smaller than 450 [MWh] is Category A and the range where the demand is larger than 450 [MWh] is Category B, the profit and loss in Category B is the profit and loss in Category A. Compared to the overall decrease. This is because when the power demand exceeds 450 [MWh], a new generator is started to meet the power demand.

Normally, generators are started in the order in which the unit price of power generation is cheaper, so when a new start occurs, the average unit price of power generation rises and profit or loss decreases. Since the aspect of the profit and loss distribution changes before and after the classification in this way, the modeling error becomes large when the relationship between the uncertain factors of the classification A and the classification B and the profit and loss is represented by one model. This leads to the possibility that the accuracy of the profit / loss omission calculation in the next step S107 deteriorates.

For this reason, the range of data used for modeling is divided according to the power generation capacity of each generator in ascending order of power generation unit price, and by modeling for each division, more accurate modeling is carried out compared to before the division. ..

For example, suppose that there are five generators owned by an electric power company, the power generation capacity of each generator is 100 [MW], and the demand to be supplied fluctuates from 300 [MW] to 500 [MW]. In this case, two categories are provided, and each of the two categories A and B is classified as category A: 300 [MW] to 400 [MW] and category B: 400 [MW] to 500 [MW], and uncertain factors and profit / loss data. Is divided into categories A and B, and model equations are created in each category as in equations (4) and (5).

Here, the subscripts a and b represent each division, and a ₀ to a ₃ are model parameters. By calculating these model parameters using data separated for each category, the relationship between uncertainties and profit and loss can be modeled for each category.

When modeling the relationship between uncertainties and profit and loss, the division model creation unit 20 sets the range of data used for modeling according to the capacity of the generator with respect to the demand that the electric power company can supply energy to. It will be divided and modeled for each division.

By modeling for each category, it is possible to avoid an increase in modeling error due to changes in the appearance of the profit and loss distribution before and after the category when no category is used, and to model the relationship between uncertain factors and profit and loss. The accuracy can be improved.

In the electric power business profit / loss analysis system according to the second embodiment, in consideration of the property that the average power generation unit price differs depending on the number and type of start-up generators required to satisfy the demand, the division model creation unit 20 uses the start-up generators. By dividing the range of data used for modeling according to the capacity and modeling for each division, it is possible to improve the accuracy of modeling regarding the relationship between uncertain factors and profit / loss.

In the second embodiment, the electric power demand has been described as an example of the demand that can be supplied with energy by the generator, but other energy may be used, and thermal energy or the like may be considered.

As described above, the electric power business profit and loss analysis system includes a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, and a scenario generation unit that generates N future forecast scenarios related to uncertain factors. , The power generation planning department that calculates M profits and losses for each scenario from the assumption of power demand and supply based on the scenario, and the profit and loss model that divides the relationship between the M scenarios and profits and losses by the capacity of the generator. The division model creation unit, the abbreviated calculation unit that calculates the profit and loss for each scenario by substituting (NM) scenarios that were not calculated by the power generation planning department into the profit and loss model, and the M that was calculated by the power generation planning department. It is equipped with a risk amount calculation unit that calculates the risk from the profit and loss of each unit and the (NM) number of profit and loss calculated by the abbreviated calculation unit, and a result display unit that displays the risk of the profit and loss calculated by the risk amount calculation unit. There is. However, N and M are natural numbers, and there is a relationship of M <N.

Embodiment 3.
In the first embodiment, the operation of the electric power business profit / loss analysis system for quickly obtaining the electric power business profit / loss in all cases generated for uncertain factors is shown. Further, in the second embodiment, the accuracy of modeling can be improved by classifying the data to be used according to the magnitude of demand and modeling for each class. In the third embodiment, a method for obtaining data for modeling is described, and the purpose is to improve the modeling accuracy as in the second embodiment.

As shown in FIG. 9, the third embodiment includes a scenario extraction unit 30 in addition to the first embodiment. In FIG. 9, the same number is assigned to the same configuration as in the first embodiment.

The scenario extraction unit 30 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the scenario extraction unit 30, and stores the model obtained as a result in at least one of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. .. Further, the processing result may be displayed on an output device 7 such as a display and shown to the operator.

FIG. 10 is a flowchart showing the operation of the electric power business profit / loss analysis system according to the third embodiment. Hereinafter, the operation will be described with reference to FIG. In FIG. 10, the same processing as in the first embodiment is assigned the same number, and the description thereof will be omitted.

In step S300 of FIG. 10, the scenario extraction unit 30 ranks the scenarios generated by the scenario generation unit 12, and extracts the scenarios of case M [pieces] used for model generation based on the ranking.

For example, FIG. 11 shows 20 fuel price scenarios as uncertain factors. The horizontal axis is the time axis at predetermined intervals, where the day and the vertical axis are the fuel price [dollar / barrel]. Assuming that the order in which the scenarios were generated is as shown in the figure, in the case of M = 5, only the scenarios in which the fuel price was relatively high during the profit and loss calculation period are the start / stop plans and outputs of S103 and S104 in the next step. There is a possibility to handle it in a decision and calculate profit and loss. If a profit / loss model is created from these scenarios and profit / loss, the profit / loss in the scenario where the fuel price falls due to the bias of the scenario data may not be accurately obtained from the profit / loss model.

Therefore, by extracting the scenarios to be used in S103 and S104 in advance in the scenario extraction unit 30, it is possible to prevent data bias and prevent deterioration of modeling accuracy.

As a scenario extraction method, for example, there is a method of calculating the average value of each scenario, ranking the average value in descending order, and extracting M scenarios from the scenario with the highest ranking to the Nth scenario at equal intervals. If there are multiple uncertain factors, the above-mentioned ranking and scenario extraction will be performed for each uncertain factor, and scenarios with the same ranking (for example, the 1st place exchange / fuel price scenario will be the 1st and 5th place exchanges). -There is a method to make the fuel scenario into a case by collecting cases 2 etc.).

The scenario extraction unit 30 ranks the generated scenarios and selects the scenario to be used by the power generation planning unit 13.

The electric power business profit / loss analysis system according to the third embodiment ranks the generated scenarios and makes selections based on the rankings to prevent bias of input data when modeling the relationship between uncertain factors and profit / loss. By doing so, the accuracy of modeling can be improved.

In the third embodiment, the electric power demand has been described as an example of the demand that can be supplied with energy by the generator, but other energy may be used, and thermal energy or the like may be considered.

The electric business profit and loss analysis system has a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, a scenario generation unit that generates N future forecast scenarios related to uncertain factors, and power based on the scenarios. The power generation planning department calculates M profits and losses for each scenario from the demand and supply assumptions, the model creation department that uses the relationship between the M scenarios and profits and losses as a profit and loss model, and the power generation planning department that calculates the profit and loss model. The abbreviated calculation unit that substitutes the missing (NM) scenarios and calculates the profit and loss for each scenario, and the M profit and loss calculated by the power generation planning unit and the (NM) number calculated by the abbreviated calculation unit. It has a risk amount calculation unit that calculates the risk from the profit and loss, and a result display unit that displays the profit and loss risk calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

In addition, it is equipped with a scenario extraction unit that extracts M scenarios in descending order of priority from the N scenarios generated by the scenario generation unit, and the power generation planning unit extracts the M scenarios extracted by the scenario extraction unit. I am using it.

Embodiment 4.
In the second embodiment, the accuracy of modeling can be improved by classifying the data to be used according to the magnitude of demand and modeling for each class. Further, in the third embodiment, the modeling accuracy is improved by devising a method for obtaining data for modeling. The fourth embodiment is a combination of the second embodiment and the third embodiment. More specifically, the model creation unit 14 is replaced with the division model creation unit 20 and the scenario extraction unit 30 is added to the first embodiment.

FIG. 12 is a block diagram showing the function of the electric power business profit / loss analysis system according to the fourth embodiment. Further, FIG. 13 is a flowchart showing the operation of the electric power business profit / loss analysis system according to the fourth embodiment. Each component and step is as described in Embodiment 1 to Embodiment 3.

In the electric power business profit / loss analysis system according to the fourth embodiment, in consideration of the property that the average power generation unit price differs depending on the number and type of start-up generators required to satisfy the demand, the division model creation unit 20 uses the start-up generators. By dividing the range of data used for modeling according to the capacity and modeling for each division, it is possible to improve the accuracy of modeling regarding the relationship between uncertain factors and profit / loss.

In addition, the electric power business profit / loss analysis system according to the fourth embodiment ranks the generated scenarios and makes selections based on the rankings, and biases the input data when modeling the relationship between uncertain factors and profit / loss. By preventing this, the accuracy of modeling can be improved.

In the fourth embodiment, the electric power demand has been described as an example of the demand that can be supplied with energy by the generator, but other energy may be used, and thermal energy or the like may be considered.

The electric business profit and loss analysis system has a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, a scenario generation unit that generates N future forecast scenarios related to uncertain factors, and power based on the scenarios. The power generation planning department that calculates M profits and losses for each scenario from the demand and supply assumptions of Omission calculation unit that calculates profit and loss for each scenario by substituting (NM) scenarios that were not calculated by the power generation planning department into the profit and loss model, and M profit and loss and omission calculated by the power generation planning department. It is provided with a risk amount calculation unit that calculates the risk from (NM) pieces of profit and loss calculated by the calculation unit, and a result display unit that displays the risk of the profit and loss calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

In addition, it is equipped with a scenario extraction unit that extracts M scenarios in descending order of priority from the N scenarios generated by the scenario generation unit, and the power generation planning unit extracts the M scenarios extracted by the scenario extraction unit. I am using it.

Embodiment 5.
In the first embodiment of the present invention, it was shown that an approximate profit / loss distribution can be obtained at high speed by performing an abbreviated calculation with respect to the profit / loss distribution when the power generation plan is calculated for all scenarios. When making a management decision, after roughly grasping the overall tendency of the profit and loss distribution, the base of the profit and loss distribution (especially the side where the loss occurs) is determined in order to determine whether or not there is a possibility of suffering a profit or loss. It is desirable to know as accurately as possible.

Therefore, in the fifth embodiment, in the scenario distribution of the profit / loss obtained by the omitted calculation unit 15 of the first embodiment as shown in FIG. On the other hand, the purpose is to recalculate in the power generation planning unit 13 and accurately grasp whether or not the profit or loss will be incurred. In FIG. 14, the horizontal axis is the profit / loss and the vertical axis is the frequency, and the distribution of the profit / loss after the abbreviated calculation is shown. There is.

As shown in FIG. 15, the fifth embodiment includes a recalculation target determination unit 50 in addition to the first embodiment. In FIG. 15, the same number is assigned to the same configuration as in the first embodiment.

The recalculation target determination unit 50 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the recalculation target determination unit 50, and sets the recalculation target scenario obtained as a result to any of the main storage device 3, the auxiliary storage device 4, and the external storage device 5. Or save in multiples. Further, the processing result may be displayed on an output device 7 such as a display and shown to the operator.

Further, in order for the recalculation target determination unit 50 to determine the recalculation target scenario, the data setting unit 10 specifies the range to be recalculated in advance, so that the profit / loss is small in the profit / loss distribution, in other words, the profit / loss becomes negative. Enter or acquire the recalculation target ratio (%) that determines how many easy scenarios to select, or the recalculation target profit / loss amount (yen) for selecting only scenarios in which the profit / loss amount is less than or equal to a predetermined value. , Save to the specified storage device.

FIG. 16 is a flowchart showing the operation of the electric power business profit / loss analysis system according to the fifth embodiment. Hereinafter, the operation will be described with reference to FIG. In FIG. 16, the same processing as in the first embodiment is assigned the same number, and the description thereof will be omitted.

In step S500 of FIG. 16, the recalculation target determination unit 50 determines the recalculation target scenario. More specifically, for example, a scenario in which the profit / loss amount is smaller than the recalculation target profit / loss amount (yen) is selected as a recalculation target scenario by the recalculation target ratio (%) from the scenario with a small profit / loss. Select as.

In step S501, the power generation planning unit 13 recalculates the scenario to be recalculated selected in step S500.

In step S502, the risk amount calculation unit 16 calculates a desired risk index from N profit and loss calculation results including the recalculated result, and stores it in the data storage unit 11.

In step S503, the result display unit 17 displays the calculation result including the recalculation result stored in the data storage unit 11.

The electric power business profit / loss analysis system according to the fifth embodiment has a power generation planning unit 13 for a part of the profit / loss distribution obtained by the abbreviated calculation, which is at the base of the profit / loss side where the profit / loss becomes small or becomes a loss. By recalculating with, it is possible to accurately grasp whether or not to suffer profit or loss.

The electric business profit and loss analysis system has a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, a scenario generation unit that generates N future forecast scenarios related to uncertain factors, and power based on the scenarios. The power generation planning department that calculates M profits and losses for each scenario from the demand and supply assumptions of Omission calculation unit that calculates profit and loss for each scenario by substituting (NM) scenarios that were not calculated by the power generation planning department into the profit and loss model, and M profit and loss and omission calculated by the power generation planning department. It is provided with a risk amount calculation unit that calculates the risk from (NM) pieces of profit and loss calculated by the calculation unit, and a result display unit that displays the risk of the profit and loss calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

In addition, the power generation planning unit recalculates the scenario selected by the recalculation target determination unit, which includes a recalculation target determination unit that selects a predetermined scenario on the loss side from the distribution of profit and loss obtained from the omission calculation unit.

Embodiment 6.
In the first embodiment of the present invention, the power generation planning unit 13 performs calculations for predetermined M scenarios before starting the abbreviated calculation. If the purpose is to finish the calculation in a predetermined time, the value of M may be set in advance as in the first embodiment. However, an appropriate value of M for efficiently reproducing the profit and loss distribution is the number of calculations performed by the power generation planning unit 13 as much as necessary.

As the power generation planning unit 13 proceeds with the calculation, the indicators such as the expected value of profit and loss and the variance calculated from the calculated scenario will converge. At this time, if these indexes are within a predetermined error range, it can be determined that they have converged, and the power generation planning unit 13 can determine the end of the calculation.

Therefore, in the sixth embodiment, the number of calculations performed by the power generation planning unit of the first embodiment is not limited, and as shown in FIG. 17, the expected profit / loss obtained by accumulating the results calculated by the power generation planning unit 13 is expected. The purpose is to end the calculation in the power generation planning unit 13 when the indexes such as the value and the dispersion have converged, and to optimize the number of calculations in the power generation planning unit 13. As a result, wasteful calculation time can be shortened.

The horizontal axis of FIG. 17 is the number of scenario calculations, and the vertical axis is the convergence test index. When the index for determining convergence converges and falls within a predetermined range (the portion surrounded by the dotted line), it is determined that the index has converged, and the calculation of the power generation planning unit 13 can be completed here.

Here, the index for determining convergence is not limited to the expected value of profit and loss and variance, but may be skewness, kurtosis, VaR, CVaR, and the like. The index used for the determination may be one of these, or a plurality of indexes may be used in combination.

Here, VaR is an abbreviation for Value at Risk, and is a concept that considers the range of profit and loss that can normally occur and regards the worst loss in that range as the amount of risk. For example, when the confidence interval is 99%, the value corresponding to 99% from the one with the best profit and loss is VaR.

CVaR is an abbreviation for Conditional Value at Risk, and is an average profit or loss when it exceeds the range of profit or loss that can normally occur. For example, when the confidence interval is 99%, the average value of the values obtained by excluding 99% from the one with the best profit and loss is CVaR.

For example, the calculated number of scenarios is n, the amount of profit / loss of scenario i (i: 1, 2, ..., N) is x (i), the average value of x (i) is x_ave, and the standard deviation is s. Then, the skewness indicating the asymmetry of the distribution can be calculated by, for example, the following equation (6).

Further, the kurtosis indicating the sharpness of the distribution can be calculated by, for example, the following equation (7).

As shown in FIG. 18, the sixth embodiment includes a convergence test unit 60 in addition to the first embodiment. In FIG. 18, the same number is assigned to the same configuration as in the first embodiment.

The convergence test unit 60 executes, for example, a program developed on the main storage device 3 shown in FIG. 2 on the CPU 2. The CPU 2 reads the data on the main storage device 3, processes the convergence judgment unit 60, and sets the convergence judgment index such as the expected value of profit / loss and the dispersion obtained as a result in the main storage device 3, the auxiliary storage device 4, and the outside. It is stored in any one or a plurality of storage devices 5. Further, the processing result may be displayed on an output device 7 such as a display and shown to the operator. Further, the convergence determination unit 60 determines that the convergence determination index such as the expected value of profit / loss and the variance is within a predetermined error range, and instructs the power generation planning unit 13 to end the calculation. do.

Further, in order to determine whether or not the convergence determination unit 60 has converged, the data setting unit 10 is used to determine in advance whether or not the convergence determination index such as the expected value of profit or loss and the variance is within a predetermined error range. The error reference ε is input or acquired and stored in the specified storage device. Using this error standard ε, for example, the absolute value of the difference between the previous value X (t-1) and the current value X (t) of a certain convergence judgment index X is less than the error standard ε, as shown in the following equation. If there is, it is judged that it has converged.

| X (t) -X (t-1) | <ε ・ ・ ・ (8)

In addition, in order to eliminate chance, it is not determined that the convergence test index has converged only by collecting one within a predetermined error range, but that it has converged when it has been settled several times in a row. You may.

Further, as shown in FIG. 17, the determination start timing (times) and the determination end timing (times) are set so that the determination is performed between a certain number of calculations (judgment start timing) and another calculation number (judgment end timing). It may be set. However, it is assumed that the judgment start timing <judgment end timing holds.

FIG. 19 is a flowchart showing the operation of the electric power business profit / loss analysis system according to the sixth embodiment. Hereinafter, the operation will be described with reference to FIG. In FIG. 19, the same processing as in the first embodiment is assigned the same number, and the description thereof will be omitted.

In step S600 of FIG. 19, in the convergence test unit 60, if it is equal to or longer than the determination start timing, the process proceeds to step S601, and if it is less than the determination start timing, the process returns to step S103. More specifically, the convergence test unit 60 determines whether or not the number of calculations in the power generation planning unit 13 is equal to or greater than the determination start timing, and if it is equal to or greater than the determination start timing, step S601 is performed. Then return to S103.

In step S601, if it is determined that the convergence determination unit 60 has converged, the process proceeds to step S106, and if it is determined that the convergence test unit 60 has not converged, the process proceeds to step S602. More specifically, each time the power generation planning unit 13 calculates each scenario, the convergence judgment index such as the expected value of profit / loss and the dispersion for all the calculated scenarios is calculated, and the convergence judgment index is the data setting unit 10. It is determined whether or not it is within the predetermined error range based on the set error standard, and if it is within the error range once or several times in a row, the power generation planning unit 13 is instructed to end the calculation, and step S106 Is executed, and if it is out of the error range, the process proceeds to step S602.

In step S602, it is determined whether the number of calculations in the power generation planning unit 13 is equal to or greater than the determination end timing, and if it is equal to or greater than the determination end timing, the power generation planning unit 13 is instructed to end the calculation, and the process proceeds to step S106. If it is equal to or less than the determination end timing, step S103 is executed.

Equipped with a recalculation target determination unit 50 that selects a predetermined scenario on the loss side from the distribution of profit and loss obtained from the M scenarios of the power generation planning unit 13 and the (NM) scenarios of the omitted calculation unit 15. ing. However, it is natural that recalculation can be omitted if the same calculation is performed for M pieces for which the power generation planning unit 13 has already calculated the profit and loss.

The electric power business profit / loss analysis system according to the sixth embodiment does not limit the number of times of calculation by the power generation planning unit 13, and the expected value, diversification, etc. of the profit / loss obtained by accumulating the calculation results by the power generation planning department 13. By ending the calculation in the power generation planning unit 13 when the index has converged, it is possible to optimize the number of calculations in the power generation planning unit 13.

It is important whether or not the profit and loss of the power generation planning unit 13 has converged, and any index for the judgment may be used. Further, the criterion for determining whether or not the product has converged is a problem in relation to the allowable calculation time, and cannot be unconditionally determined.

The electric business profit and loss analysis system has a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, a scenario generation unit that generates N future forecast scenarios related to uncertain factors, and power based on the scenarios. The power generation planning department that calculates M profits and losses for each scenario from the demand and supply assumptions of Omission calculation unit that calculates profit and loss for each scenario by substituting (NM) scenarios that were not calculated by the power generation planning department into the profit and loss model, and M profit and loss and omission calculated by the power generation planning department. It is provided with a risk amount calculation unit that calculates the risk from (NM) pieces of profit and loss calculated by the calculation unit, and a result display unit that displays the risk of the profit and loss calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

The power generation planning unit includes a convergence determination unit that determines whether or not the profit and loss of the power generation planning unit has converged, and the power generation planning unit ends the calculation when the convergence determination unit determines that the convergence has occurred.

Embodiment 7.
In the fifth embodiment of the present invention, in order to accurately grasp the base of the profit / loss distribution, the power generation planning unit 13 recalculates some scenarios in the base of the side where the profit / loss becomes small or becomes a loss. Showed to do. In order to grasp the base of the profit and loss distribution more accurately, it is necessary to generate many scenarios existing at the base of the profit and loss distribution, increase the number of samples of the base of the profit and loss distribution, and express the shape of the base in more detail. ..

Therefore, in the seventh embodiment, the scenario generation unit 12 of the first embodiment generates a scenario that is likely to exist at the base of the profit / loss distribution, and expresses the shape of the base of the profit / loss distribution more finely and accurately. The purpose is to do.

In order to realize this, a feature extraction unit 70 as shown in FIG. 20 is provided, and the feature extraction unit 70 substitutes a scenario generated based on a future forecast regarding uncertain factors into a profit / loss model, and as a result, the base of the profit / loss distribution. Focus on the scenarios that exist in, and extract the characteristics of those scenarios. Uncertainties that characterize the scenario include, for example, high and low exchange rates, high and low fuel prices, and high and low demand. From these characteristics, when the scenario generation unit 12 generates a scenario, the range of possible values of uncertain factors is narrowed, and a scenario that is likely to exist at the base of the profit and loss distribution is generated.

By doing so, for example, when there is a high possibility that a scenario with high demand exists at the base of the profit and loss distribution, only the scenario with high demand is generated, and the number of samples at the base of the profit and loss distribution is increased. , It is possible to express the shape of the base more finely and accurately. Not limited to one uncertain factor, it may be characterized by a combination of multiple uncertain factors. For example, there is a high possibility that a scenario characterized by high exchange rate and high demand exists at the base of the profit and loss distribution. In some cases, only scenarios with high exchange rates and high demand may be generated.

The electric business profit / loss analysis system according to the seventh embodiment extracts the characteristics of the scenario existing at the base of the profit / loss distribution, and the scenario generation unit 12 generates the scenario based on the extracted characteristics to generate the base of the profit / loss distribution. It is possible to increase the number of samples and express the shape of the base of the profit and loss distribution more finely and accurately.

The electric business profit and loss analysis system has a data setting unit that sets data including uncertain factors related to the risk of profit and loss of the electric business, a scenario generation unit that generates N future forecast scenarios related to uncertain factors, and power based on the scenarios. The power generation planning department that calculates M profits and losses for each scenario from the demand and supply assumptions of Omission calculation unit that calculates profit and loss for each scenario by substituting (NM) scenarios that were not calculated by the power generation planning department into the profit and loss model, and M profit and loss and omission calculated by the power generation planning department. It is provided with a risk amount calculation unit that calculates the risk from (NM) pieces of profit and loss calculated by the calculation unit, and a result display unit that displays the risk of the profit and loss calculated by the risk amount calculation unit. However, N and M are natural numbers, and there is a relationship of M <N.

Further, a feature extraction unit for extracting the characteristics of the scenario from the profit and loss obtained from the power generation planning unit and the abbreviated calculation unit is provided, and the scenario generation unit generates a scenario having the characteristics extracted by the feature extraction unit.

The invention of the present application is not limited to the embodiments described so far, and can be variously modified within the scope of the invention of the present application. That is, the configuration of the embodiments described so far may be appropriately improved, or at least a part thereof may be replaced with another configuration. Further, the configuration requirement without particular limitation on the arrangement is not limited to the arrangement disclosed in the embodiment, and can be arranged at a position where the function can be achieved. In addition, the invention may be formed by appropriately combining a plurality of components disclosed in the embodiments described so far. Furthermore, the invention of the present application is shown by the scope of claims, not the scope of the embodiments described so far, and includes all modifications within the meaning and scope equivalent to the scope of claims.

1 Computer, 2 CPU, 3 Main memory, 4 Auxiliary storage, 5 External storage, 6 Network, 7 Output, 8 Input, 10 Data setting, 11 Data storage, 12 Scenario generator, 13 Power generation plan Unit, 14 Model creation unit, 15 Omission calculation unit, 16 Risk amount calculation unit, 17 Result display unit, 20 Division model creation unit, 30 Scenario extraction unit, 50 Recalculation target determination unit, 60 Convergence judgment unit, 70 Feature extraction unit ..

Claims (7)

A data setting unit that sets data including uncertainties regarding the risk of profit and loss in the electric power business,
A scenario generator that generates N future prediction scenarios for the uncertain factors,
A power generation planning unit that calculates M profits and losses for each scenario from the assumption of power demand and supply based on the scenario.
A model creation unit that uses the relationship between the M scenarios and the profit and loss as a profit and loss model,
An abbreviated calculation unit that calculates the profit and loss for each scenario by substituting (NM) the scenarios that were not calculated by the power generation planning unit into the profit and loss model.
A risk amount calculation unit that calculates the risk from the M profits and losses calculated by the power generation planning unit and the (NM) profits and losses calculated by the abbreviated calculation unit.
An electric utility profit / loss analysis system including a result display unit for displaying the risk of the profit / loss calculated by the risk amount calculation unit.
However, N and M are natural numbers, and there is a relationship of M <N. A data setting unit that sets data including uncertainties regarding the risk of profit and loss in the electric power business,
A scenario generator that generates N future prediction scenarios for the uncertain factors,
A power generation planning unit that calculates M profits and losses for each scenario from the assumption of power demand and supply based on the scenario.
A division model creation unit that converts the relationship between the M scenarios and the profit / loss into a profit / loss model in which the demand for the electric power is divided by the capacity of the generator.
An abbreviated calculation unit that calculates the profit and loss for each scenario by substituting (NM) the scenarios that were not calculated by the power generation planning unit into the profit and loss model.
A risk amount calculation unit that calculates the risk from the M profits and losses calculated by the power generation planning unit and the (NM) profits and losses calculated by the abbreviated calculation unit.
An electric utility profit / loss analysis system including a result display unit for displaying the risk of the profit / loss calculated by the risk amount calculation unit.
However, N and M are natural numbers, and there is a relationship of M <N. The electric utility profit / loss analysis system according to claim 1 or 2.
It is provided with a scenario extraction unit that extracts the M scenarios in descending order of priority from the N scenarios generated by the scenario generation unit.
The power generation planning unit is an electric power business profit / loss analysis system characterized by using the M scenarios extracted by the scenario extraction unit. The electric utility profit / loss analysis system according to any one of claims 1 to 3.
The profit and loss model is an electric utility profit and loss analysis system characterized by using Monte Carlo simulation. The electric utility profit / loss analysis system according to any one of claims 1 to 4.
It is provided with a recalculation target determination unit that selects a predetermined scenario on the loss side from the distribution of profit and loss obtained from the omitted calculation unit.
The power generation planning unit is an electric power business profit / loss analysis system characterized in that the scenario selected by the recalculation target determination unit is recalculated. The electric utility profit / loss analysis system according to any one of claims 1 to 5.
A convergence test unit for determining whether or not the profit and loss of the power generation planning unit has converged is provided.
The power generation planning unit is an electric power business profit / loss analysis system characterized in that the calculation is completed when the convergence determination unit determines that the convergence is achieved. The electric utility profit / loss analysis system according to any one of claims 1 to 6.
It is provided with a feature extraction unit that extracts the characteristics of the scenario from the profit and loss obtained from the power generation planning unit and the abbreviated calculation unit.
The scenario generation unit is an electric utility profit / loss analysis system characterized by generating a scenario having features extracted by the feature extraction unit.

Images