Prediction method and simulation device

Abstract

To provide a prediction method and a simulation device that reduce the running cost of a power supply system having a main power supply and an auxiliary power supply.SOLUTION: Through a random number calculation assuming that a prediction error of a first power amount y supplied to a user from a main power supply which generates electric power with renewable energy follows a probability distribution, a power supply system 1 generates a plurality of combinations of an estimated value of the first power amount and an estimated value of a second power amount x(r) that is supplied to the user from an auxiliary power supply that can adjust power to be supplied, and calculates a first cost required for generating power of the estimated value of the second power amount with the auxiliary power supply, and a second cost indicating a penalty for the case where a total power amount, which is a sum of a value that can be taken by the first power amount for a predicted demand power amount and the estimated value of the second power amount, is insufficient. On the basis of the estimated value of the second power amount associated with a combination that minimizes an expected value of the running cost, which increases in accordance with the sum of the first and second costs, parameter values are calculated.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a prediction method and a simulation device. [Background technology]

[0002] In order to realize a sustainable society, there is a demand for the use of renewable energy such as photovoltaic power generation (PV). However, since the amount of power generated by photovoltaic power generation fluctuates greatly, it is necessary to compensate for the instability of the output. Therefore, as disclosed in Patent Document 1, a power supply system has been proposed that combines photovoltaic power generation with, for example, a fuel cell to supply power commensurate with the amount of power demanded by consumers, and compensates for the power shortage of photovoltaic power generation with the power of the fuel cell. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2002-075388 A Summary of the Invention [Problem to be solved by the invention]

[0004] Considering the running costs of power supply, it is desirable to maximize the amount of power supplied from the main power source (photovoltaic power generation) and minimize the amount of power supplied from the auxiliary power source (fuel cell). Also, to keep running costs down, it is desirable to keep the amount of fuel used by the auxiliary power source to a minimum. In this context, it is necessary to estimate in advance the amount of power supplied from the auxiliary power source in order to appropriately reduce running costs.

[0005] The present invention has been made in light of the above-mentioned circumstances, and has an object to provide a prediction method and a simulation device that can reduce the running costs of a power supply system equipped with a main power supply and an auxiliary power supply. [Means for solving the problem]

[0006] In order to achieve the above object, a prediction method according to a first aspect of the present invention comprises: A prediction method for predicting, by a simulation device, a parameter for determining a first amount of power supplied to a consumer from a main power source that generates power using renewable energy and a second amount of power supplied to the consumer from an auxiliary power source capable of adjusting the power supply, such that a total amount of power is equal to a predicted demand amount of power per unit time of the consumer, generating a plurality of combinations of the estimated value of the first amount of power and the estimated value of the second amount of power based on an initial predicted value of the first amount of power and the predicted amount of power demand by random number calculation assuming that a prediction error of the first amount of power follows a probability distribution; calculating a first cost required for generating the estimated value of the second amount of power by the auxiliary power source for each of the combinations; calculating, for each of the combinations, a plurality of possible values ​​of the first amount of power based on the estimated value of the first amount of power by random number calculation assuming that deviation from the estimated value of the first amount of power follows the probability distribution, and calculating, for each possible value of the first amount of power, a second cost indicating a penalty when the total amount of power is insufficient for the predicted amount of demanded power; The value of the parameter is calculated based on the estimated value of the second amount of power associated with the combination that minimizes an expected value of a running cost that increases according to the sum of the first cost and the second cost.

[0007] calculating the first cost based on a unit price of fuel used to supply power to the auxiliary power source, a power generation efficiency of the auxiliary power source, and an estimated value of the second amount of power; This may also be the case.

[0008] When the total amount of power based on a possible value of the first amount of power is insufficient with respect to the predicted amount of power demand, a value proportional to the amount of the shortage is calculated as the second cost; When the total amount of power based on a possible value of the first amount of power is not insufficient for the predicted amount of demanded power, the second cost is set to 0. This may also be the case.

[0009] the auxiliary power source is a fuel cell, and the parameter is a load factor of the fuel cell; calculating a value of the parameter based on an estimated value of the second amount of power associated with the combination that minimizes an expected value of a running cost that increases in accordance with the sum of the first cost, the second cost, and a third cost that increases in accordance with an increase in the load factor; This may also be the case.

[0010] Calculating a load factor of the fuel cell for each time period having a predetermined time span; If the load factor of the time period of interest has increased compared to the load factor of the time period immediately preceding it, a value proportional to the amount of increase is calculated as the third cost; When the load factor of the time period of interest has not increased compared to the previous time period, the third cost is set to 0. This may also be the case.

[0011] The main power source is a solar power generation device. This may also be the case.

[0012] A simulation device according to a second aspect of the present invention comprises: A simulation device that predicts a parameter that determines a first amount of power supplied to a consumer from a main power source that generates power using renewable energy and a second amount of power supplied to the consumer from an auxiliary power source that can adjust the power supply, so that a total amount of power is equal to a predicted amount of power demand per unit time of the consumer, the simulation device comprising: an estimate calculation unit that generates a plurality of combinations of an estimated value of the first amount of power and an estimated value of the second amount of power based on an initial predicted value of the first amount of power and the predicted amount of power demand by random number calculation assuming that a prediction error of the first amount of power follows a probability distribution; a first cost calculation unit that calculates a first cost required for generating the estimated value of the second amount of power by the auxiliary power source for each of the combinations; a second cost calculation unit that calculates, for each of the combinations, a plurality of possible values ​​of the first amount of power based on the estimated value of the first amount of power by random number calculation assuming that a deviation from the estimated value of the first amount of power follows the probability distribution, and calculates, for each possible value of the first amount of power, a second cost that indicates a penalty when the total amount of power is insufficient for the predicted amount of power demand; a parameter calculation unit that calculates a value of the parameter based on an estimated value of the second amount of power associated with the combination that minimizes an expected value of a running cost that increases according to a sum of the first cost and the second cost; Equipped with. Effect of the Invention

[0013] According to the present invention, it is possible to estimate the running cost while taking into account the prediction error by assuming that the prediction error of the first amount of power supplied from the main power source follows a probability distribution, and to calculate the parameter values ​​that minimize the expected value of the running cost, thereby reducing the running cost of the power supply system equipped with the main power source and the auxiliary power source. [Brief description of the drawings]

[0014] [Figure 1] 1 is a block diagram showing a configuration of a power supply system according to an embodiment of the present invention; [Diagram 2] 1 is a bar graph showing a relationship between a first amount of power supplied from a main power source, a second amount of power supplied from an auxiliary power source, and a predicted amount of power demand of a consumer. [Diagram 3] 1 is a block diagram showing a functional configuration of a simulation device according to an embodiment of the present invention. [Figure 4] 4 is a graph showing an approximation formula indicating the relationship between the load factor and power generation efficiency in an auxiliary power supply. [Diagram 5] FIG. 13 is a schematic diagram showing a calculation flow. [Figure 6]FIG. 2 is a block diagram showing a hardware configuration of the simulation device. [Figure 7] 1 is a flowchart of a prediction method according to an embodiment of the present invention. [Figure 8] (A) is an example of a histogram of running costs calculated by Method I. (B) is an example of a histogram of running costs calculated by Method II. [Figure 9] (A) is a graph showing the change in load factor in multiple time zones in one day. (B) is a graph showing the change in expected value of running cost in multiple time zones in one day. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In each drawing, the same or equivalent parts are denoted by the same reference numerals.

[0016] In this embodiment, the prediction is performed on a power supply system 1 shown in Fig. 1. As shown in Fig. 1, the power supply system 1 includes a main power source 2 and an auxiliary power source 3, and supplies power to a consumer 4.

[0017] The main power source 2 is a power source that generates power using renewable energy and supplies the generated power to the outside. In this embodiment, the main power source 2 is a solar power generation device that generates solar power. The main power source 2 supplies the generated power to a consumer 4. The amount of power per unit time supplied from the main power source 2 to the consumer 4 is the first amount of power y. If the unit time is one hour, the first amount of power y is electric power. The magnitude of the first amount of power y varies depending on meteorological conditions such as the weather at the time.

[0018] The auxiliary power supply 3 is a power supply capable of adjusting the power it supplies. In this embodiment, the auxiliary power supply 3 is a solid oxide fuel cell (SOFC). The SOFC operates at approximately 700°C to 1000°C, and has a power generation efficiency of between 45% and 65%. The power generation efficiency of the SOFC is higher than that of other independent power generation devices. In the same system, the power generation efficiency of the SOFC is high when it operates at a high temperature, and conversely, the power generation efficiency of the SOFC is low when it operates at a low temperature.

[0019] The SOFC can adjust the power it supplies by changing the load factor r. The load factor r is the ratio of the actual output to the rated output. The rated output of the auxiliary power supply 3 is P rated Rated output P rated is the maximum amount of work per unit time in the auxiliary power supply 3. The load factor r is the rated output P rated This shows how much the auxiliary power supply 3 is operating relative to the rated output P rated The load factor when the auxiliary power supply 3 performs a workload of x is expressed by the following formula. r=x / P rated …(1) This x corresponds to x(r), and the amount of power generated (output) per unit time is x(r) = r P rated It becomes.

[0020] In this embodiment, the SOFC is operated at a partial load. Partial load operation means power generation with a surplus of power generation capacity, for example, operating an SOFC with a rated output of 1000 W at an output of 500 W or 300 W.

[0021] In partial load operation, the power generation efficiency of the SOFC (system efficiency including energy loss due to auxiliary equipment) is increased by applying a load to the SOFC (increasing the current flowing therethrough). Since the SOFC has excellent load responsiveness, the power generation efficiency of the SOFC increases rapidly when a load is applied to the SOFC. This power supply system 1 is a power generation system that utilizes the load responsiveness of the SOFC in partial load operation, and achieves a stable power supply by changing the load factor r of the SOFC according to the first power amount y of solar power generation.

[0022] The auxiliary power source 3 supplies power corresponding to a load factor r to the consumer 4. The amount of power per unit time supplied from the auxiliary power source 3 to the consumer 4 is the second amount of power x(r). In this embodiment, the parameter that determines the second amount of power x(r) is the load factor r.

[0023] The power supplied from the main power source 2 and the auxiliary power source 3 is consumed by a consumer 4. The amount of power demanded by the consumer 4 per unit time is assumed to be f. Ideally, the following relationship should basically hold between the amount of power demanded by the consumer 4, f, the first amount of power y supplied from the main power source 2, and the second amount of power x(r) supplied from the auxiliary power source 3. x(r)+y=f …(2) The prediction method according to the present embodiment predicts a parameter (load factor) r that determines the second amount of energy x(r) so that the total amount of energy, which is the sum of the first amount of energy y and the second amount of energy x(r), is equal to the predicted demand amount of energy f.

[0024] The power supply system 1 further includes an adjustment device 5. The adjustment device 5 adjusts the load factor r of the auxiliary power supply 3 according to the load factor r predicted by the prediction method according to the present embodiment so that the above (2) is satisfied, thereby controlling partial load operation of the SOFC. The load factor r of the auxiliary power supply 3 is adjusted by changing the current flowing through the auxiliary power supply 3. Increasing the current reduces the load factor r, and decreasing the current increases the load factor r. In this power supply system 1, the adjustment device 5 adjusts the load factor r to change the magnitude of the second amount of power x(r), thereby adjusting the load factor r so that the above (1) is satisfied.

[0025] The load factor r is adjusted by the adjusting device 5 on a time zone basis. Basically, as shown in Fig. 2, the adjustment is performed in each time zone t so that the relationship in formula (2) above is satisfied. Here, t represents the order of time zones having a predetermined time span, and the time zone immediately preceding time zone t is t-1.

[0026] In order to reduce the running cost of the power supply system 1, the simulation device 10 shown in FIG. 3 calculates in advance an estimated value y of the first amount of power y that satisfies the relationship of the above formula (2) in the time zone t. t and the estimated value x(r t ) and estimate the second power x(r t ) is determined by the parameter (load factor r t ) is calculated. t ) is calculated for multiple consecutive time zones t. In the calculation, the current parameters (load factor r t The time zone of interest for which t is being calculated is t, and the time zone immediately preceding it is t-1.

[0027] 3, the simulation device 10 is an information processing device including a calculation unit 11 and a storage unit 12. The calculation unit 11 performs various calculations. The storage unit 12 stores various information used in the calculations of the calculation unit 11.

[0028] The first power amount y in the time zone t is predicted based on the weather conditions predicted for that time period, and the value is the initial predicted value yh of the first power amount y. t The power demand f of the consumer 4 in the time zone t is input to the calculation unit 11. t The power demand f is also predicted based on the past power demand in the same time period, and is input to the calculation unit 11 as a set value. t is assumed to be the predicted power demand (predicted power demand per unit time).

[0029] The calculation unit 11 calculates an initial predicted value yh of a first power amount y in a certain time zone t. t and the predicted power demand f t Based on this, the estimated value x(r t ) and estimate the parameter (load factor) r t Calculate the load factor r t The estimate of rh t Then, the following formula holds: rh t =(f t -yh t ) / P rated …(3) Rated output P rated is a value determined based on the performance of auxiliary power supply 3, and this value is stored in storage unit 12.

[0030] The calculation unit 11 calculates a parameter (load factor) r that determines the second amount of power x(r) in the time zone t so as to reduce the running cost of the power supply system 1. t The running costs of the power supply system 1 include the following three costs.

[0031] (A) Power generation cost required to generate power using auxiliary power source 3 (first cost) Load factor r of SOFC t The power generation efficiency corresponding to η(r t ) in time zone t. The estimated value x(r t The first cost Co(r t ) is expressed as follows: Co(r t )=(U fuel / η(r t ))·x(r t )=(U fuel / η(r t ))·r t P rated …(4) Here, U fuel represents the unit cost of fuel to drive the SOFC. Also, η(rt ) is the η corresponding to the rated power rated and the load factor r t It is approximated by the following equation using and coefficient k. η(r t )=(1-(1 / (k e rt ))) / (1-1 / (k e)) η rated …(5) This approximation is obtained by fitting the experimental results with an exponential function. t Increasing the load factor r (increasing the current) reduces the power generation efficiency of the SOFC itself. However, in reality, the load factor r t Increasing the overall power generation efficiency of SOFCs η(r t ) increases. Figure 4 shows the relationship between the power generation efficiency of SOFC and the load factor, η rated A graph showing the approximate equation for the power generation efficiency η(r) when r = 0.55 and k = 3 is shown. According to this graph, the power generation efficiency η(r) is determined when the load factor r is determined. As shown in equation (5), r t =0 and η(r t ) approaches 0 very quickly, so this approximation is t = 0 near η(r t ) cannot be maintained. However, η(r t Of the three costs, the first cost Co(r t ) and the first cost Co(r t The calculation formula for r t Since there is a term, r t When is small, η(r t ), the first cost Co(r t ) itself approaches 0, so the running cost of SOFC, C t It does not affect the calculation of r t Even if is small, this approximation can be used.

[0032] As shown in FIG. 3, the memory unit 12 stores the unit price U fuel , the power generation efficiency η corresponding to the rated output of auxiliary power supply 3 rated, the coefficient k is stored. The calculation unit 11 calculates the unit price U of the fuel stored in the storage unit 12 fuel , the power generation efficiency η corresponding to the rated output of the auxiliary power supply 3 rated , based on the coefficient k and the estimated value x(r) of the second power consumption x(r t ), calculates the above formulas (4) and (5) to calculate the first cost Co(r t ).

[0033] (B) Cost (second cost) indicating the penalty when the total power consumption is insufficient with respect to the predicted demand power consumption Let g be the sum of the first power consumption y of the main power supply 2 and the second power consumption x(r) of the auxiliary power supply 3 as shown in the following formula. g = x(r) + y …(6) At this time, assume that if g < f, a penalty proportional to f - g occurs. For example, such a penalty occurs when using the Japan Electric Power Exchange (JEPX). Let the penalty unit price be U penalty , then the second cost Cp(y t ) indicating the penalty in the time zone t is expressed by the following formula. Cp(y t ) = δ(x(r t ) + y t < f t ) · U penalty · (f t - x(r t ) - y t ) …(7) Here, δ(conditional expression) is a function that becomes 1 when the conditional expression is true and 0 when it is false. As shown in Figure 3, U penalty is stored in the storage unit 12. The calculation unit 11 calculates the second cost Cp(y) shown in the above formula (7) using the U penalty stored in the storage unit 12.

[0034] (C) Cost (third cost) corresponding to the increase in the load factor of the SOFC When increasing the load factor r of the SOFC to r’ (r < r’), a third cost that increases in response to the increase in the load factor r occurs. The third cost is proportional to r’ - r, and the cost when decreasing the load factor r is 0. Therefore, the third cost Cc(r’) required when increasing the load factor from r to r’ can be expressed by the following equation using a constant α. Cc(r’) = δ(r < r’) · α · (r’ - r) …(8) Note that in the calculation in time zone t, r’ = r t resulting in r = r t-1 As shown in Figure 3, α is stored in the storage unit 12. The calculation unit 11 calculates the equation (8) using U penalty stored in the storage unit 12 to calculate the third cost Cc(r t ).

[0035] The calculation unit 11 sums up the first cost Co(r t ) calculated in (A) above, the second cost Cp(y t ) calculated in (B), and the third cost Cc(r t ) calculated in (C) to calculate the running cost C t of the SOFC-PV system 1 in time zone t. C t = Co(r t ) + Cp(y t ) + Cc(r t ) = (U fuel / η(r t )) · x(r t ) + δ(x(r t ) + y t < f) · U penalty · (f - x(r t ) - y t ) + δ(r t-1 < r t ) · α · (r t - r t-1 ) …(9) In this way, the running cost C t is composed of the first cost Co(r t ), the second cost Cp(y), and the third cost Cc(rt ) is larger than the sum of

[0036] The simulation device 10 calculates an initial predicted value yh of a first power amount y in a time zone t. t and the predicted power demand f t Based on this, the running cost C in that time zone t is calculated. t Calculate the running cost C t Based on this, the estimated value x(r t ) Parameter (load factor) r t Calculate.

[0037] Basically, the initial prediction value yh of the first power amount y t and the predicted power demand f t Based on the above, the parameter (load factor) r can be calculated using the above equations (2) and (1). t In this way, the parameter (load factor) r t The method for calculating is called Method I.

[0038] However, in reality, the initial predicted value yh of the first power amount y t is the actual value y of the first power quantity y t Forecast error (y t -yh t 3, the calculation unit 11 of the simulation device 10 has an initial predicted value yh t and the probability distribution D t and the load factor r of the previous time zone t-1 t-1 Based on this, the prediction error (y t -yh t ) is the probability distribution D t The parameter (load factor) r is calculated by random number calculation assuming that t The calculation unit 11 calculates the running cost C t The function C of the parameter (load factor) r t =C(r t ) and the running cost C(r tThe calculation unit 11 calculates the running cost C(r t ) expectation E[C(r t )], the estimated value x(r t ) and the estimated value y of the first power amount y t The combination of these parameters (load factor) r t In this way, the parameter (load factor) r t The method for determining is referred to as Method II.

[0039] 3, the calculation unit 11 performs calculations according to Method II. The calculation unit 11 includes an estimated value calculation unit 20, a first cost calculation unit 21, a second cost calculation unit 22, a third cost calculation unit 23, and a parameter calculation unit 24 as components for performing this calculation.

[0040] The estimated value calculation unit 20 calculates a prediction error y t -yh t is the probability distribution D t The random number calculation is performed assuming that the probability distribution D t The estimated values ​​y of the first power quantities y distributed according to t The estimated value calculation unit 20 generates the estimated value y t For each, the initial predicted value yh of the first power amount y t and forecasted power demand f t Based on the above, an estimate y of the first power amount y is calculated. t and the estimated value x(r t Specifically, multiple combinations of the predicted demand energy f t Estimated value y of the first power amount y t By subtracting t ) to generate the above combination. In Figure 5, the estimated value (y t , x(r t The combinations of D are connected by thick vertical solid lines. t For example, the normal distribution N(μ, σ 2) can be used. μ is the mean and σ is the standard deviation. The probability distribution D t may be another distribution. As a calculation using random numbers, for example, a Monte Carlo calculation may be used. However, a calculation other than the Monte Carlo calculation may be used.

[0041] Returning to FIG. 3, the first cost calculation unit 21 calculates the estimated value y of the first power amount y t and the estimated value x(r t ) for each combination of the second power amount x(r) and t ) required to generate power from the auxiliary power source 3, t ) using the above formulas (4) and (5). First, the first cost calculation unit 21 calculates the estimated value x(r t ) and rated output Prated Based on this, the load factor r t Furthermore, the first cost calculation unit 21 calculates the η rated , coefficient k, and load factor r t Based on this, the power generation efficiency η(r t Furthermore, the first cost calculation unit 21 calculates η rated , U fuel , the estimated value x(r t ) based on the first cost Co(r t As shown in FIG. 5, the estimated value y t and the estimated value x(r t ) for each combination of parameters (load factor) r t is calculated, and the first cost Co(r t ) is calculated.

[0042] Returning to FIG. 3, the second cost calculation unit 22 calculates the estimated value y of the first power amount y t and the estimated value x(r t ) for each combination of the first power amount y t The deviation from the probability distribution D tThe estimated value y of the first power amount y is calculated by random number calculation assuming that the t Based on this, the first power amount y can have a value yb t As shown in Figure 5, the estimated value y t Probability distribution D centered on t According to the estimation, y t The first power amount y can take a value yb t The second cost calculation unit 22 generates a plurality of values ​​yb t For each day, the predicted power demand f is calculated using the above formula (7). t A second cost Cp(y) is calculated, which indicates a penalty when the total amount of power g is insufficient for the power consumption of the power supply y in the formula (7). t As, y b As shown in the formula (7), the second cost calculation unit 22 calculates the possible value yb t Total power based on g t (g) of Equation (6) in time zone t is the predicted energy demand f t If the first power amount y is insufficient, the second cost calculation unit 22 calculates a value proportional to the insufficiency as the second cost Cp(y). t Total power based on g t is the predicted power demand f t If there is no shortage for y, the second cost Cp(y) is set to 0.

[0043] Returning to FIG. 3, the third cost calculation unit 23 calculates the load factor r t The third cost Cc(r'), which increases with an increase in t Let r be r t-1 As the third cost Cc(r t As shown in the formula (9), the third cost calculation unit 23 calculates the load factor r t is the load factor r of the previous time zone t-1 t-1 If the cost is increased compared to the amount of increase, the third cost Cc(r tOn the other hand, the third cost calculation unit 23 calculates the load factor r t is the load factor r of the previous time zone t-1 t-1 If the third cost Cc(r t ) is set to 0.

[0044] Returning to FIG. 3, the parameter calculation unit 24 calculates the first cost Co(r t ) and the second cost Cp(y t ) and the third cost Cc(r t ) is the running cost C(r t The parameter calculation unit 24 calculates the possible value yb t For each, the running cost C(r t The parameter calculation unit 24 calculates the combination of the estimated values ​​(y t , x(r t )) for each running cost C(r t ) distribution, and the running cost C(r t ) expectation E[C(r t The parameter calculation unit 24 calculates the combination of the estimated values ​​(y t , x(r t )) from the running cost C(r t ) expectation E[C(r t The estimated value x(r t ) based on the value of the parameter r t Calculate.

[0045] Running cost C(r t ) includes the second cost Cp(y), which indicates the penalty when the power supply is insufficient. This value is calculated by multiplying the actual value y of the first power amount y in time zone t by y. t However, at the time of making the prediction, y t is an unknown value. Therefore, the second cost calculation unit 22 calculates the prediction error y-yh t is the probability distribution D tTherefore, in this calculation, the second cost Cp(y) is calculated by assuming that the cost varies according to the combination of the estimated values ​​(y t , x(r t )), the distribution of the second cost Cp(y) according to its variance is calculated, and the combination of the calculated estimates (y t , x(r t )) running cost C(r t ) expectation E[C(r t )] is obtained. Then, the combination of the estimated values ​​(y t , x(r t )) for each expected value E[C(r t )] is the estimated value y of the first power amount y that is the smallest. t The second power quantity x(r t ) is calculated, and the second power x(r t ) is determined by the parameter (load factor r t ) is calculated.

[0046] [Hardware configuration of the simulation device] The simulation device 10 shown in Fig. 3 is realized, for example, by a computer having the hardware configuration shown in Fig. 6 executing a software program. Specifically, the simulation device 10 includes a CPU (Central Processing Unit) 31 that controls the entire device, a main memory 32 that operates as a work area or the like for the CPU 31, an external memory 33 that stores the operating program or the like for the CPU 31, an operation unit 34, a display unit 35, a communication unit 36, and an internal bus 38 that connects these together.

[0047] The main memory 32 is composed of a RAM (Random Access Memory) etc. A program 39 to be executed by the CPU 31 is loaded into the main memory 32 from an external memory 33. The main memory 32 is also used as a working area (temporary data storage area) of the CPU 31.

[0048] The external memory 33 is composed of a non-volatile memory such as a flash memory, a hard disk, etc. The external memory 33 stores in advance a program 39 to be executed by the CPU 31.

[0049] The operation unit 34 is composed of devices such as a keyboard and a mouse, and an interface device that connects these devices to an internal bus 38 .

[0050] The display unit 35 is composed of a display device such as a CRT (Cathode Ray Tube) or a liquid crystal monitor.

[0051] The communication unit 36 ​​is an interface for data communication. Data communication with external devices is possible via the communication unit 36. For example, via this communication unit 36, setting values ​​that are prerequisites for a simulation are input from other communication destination devices, and data that is the result of a simulation is output to other communication destination devices.

[0052] The functions of the simulation device 10 can be implemented in a computer system consisting of one or more computers including one or more processors and one or more storage devices including a non-transitory storage medium. The multiple computers realize the functions of the simulation device 10 while communicating with each other via a communication network. For example, some of the multiple functions of the simulation device 10 may be implemented in one computer, and the other parts may be implemented in another computer.

[0053] Next, the operation of the simulation apparatus 10 according to this embodiment, that is, the prediction method realized by the simulation apparatus 10, will be described.

[0054] 7, first, the calculation unit 11 of the simulation device 10 inputs a set value (step S1). As a result, as shown in FIG. t , forecasted energy demand per unit time f t , probability distribution D t , the load factor r of the previous time zone t-1t-1 is input to the calculation unit 11. When performing a simulation for one time zone t, the load factor r t-1 When simulating multiple consecutive time zones t, the load factor r t-1 The load factor r of the previous time zone t-1 calculated by the calculation unit 11 is t-1 Here, various information to be stored in the storage unit 12 may be input and stored in the storage unit 12.

[0055] Next, the estimation value calculation unit 20 of the calculation unit 11 calculates the prediction error y t -yh t is the probability distribution D t As shown in FIG. 5, the estimated values ​​y of the first power amounts y are calculated by random number calculation assuming that the t is generated (step S2).

[0056] Next, the estimate value calculation unit 20 of the calculation unit 11 calculates the generated estimate value y of the first power amount y as shown in FIG. t For each time, the estimated value x(r t ) is calculated (step S3).

[0057] Next, the first cost calculation unit 21 of the calculation unit 11 calculates the estimated value x(r t ) for each load factor r t (Step S4). In the above formula (1), x = x(r).

[0058] Next, the first cost calculation unit 21 of the calculation unit 11 calculates η rated , k, and the respective load factor r t Based on this, the combination of the estimated values ​​(y t , x(r t For each of the above, the power generation efficiency η(r) is calculated (step S5).

[0059] Next, the first cost calculation unit 21 of the calculation unit 11 calculates Ufuel , each η(r), each estimate x(r t ), the combination of the estimated values ​​(y t , x(r t For each of the first and second nodes, a first cost Co(r) is calculated (step S6).

[0060] Next, the second cost calculation unit 22 of the calculation unit 11 calculates the penalty unit price U penalty , forecasted power demand f t , the combination of estimates (y t , x(r t )) based on the combination of estimates (y t , x(r t ) for each second cost Cp(y t ) is calculated (step S7). t Since is unknown at the time of simulation, it is assumed that t The first power amount can take multiple values ​​yb t Find each yb t The second cost Cp(y t ) is calculated.

[0061] Next, the third cost calculation unit 23 of the calculation unit 11 calculates a combination of the estimated values ​​in the time zone t (y t , x(r t Load factor r for each t and the load factor r of the previous time zone t-1 t-1 Based on this, the combination of the estimated values ​​(y t , x(r t )) the third cost Cc(r t ) is calculated (step S8).

[0062] Next, the parameter calculation unit 24 of the calculation unit 11 sums up the multiple Cc(r), Co(r), and Cp(y) with variations to obtain the running cost C t Calculate the combination of the estimated values ​​(y t , x(r t )) for each running cost C(rt ) expectation E[C(r t )] is calculated (step S9). As a result, the running cost C(r t ) distribution of the estimated combination (y t , x(r t )) the expected value E[C(r t )] is obtained.

[0063] Next, the parameter calculation unit 24 of the calculation unit 11 calculates the running cost C(r t ) expectation E[C(r t The combination of estimates (y t , x(r t )) based on the parameter (load factor) r in time zone t t is calculated (step S10).

[0064] When performing a simulation for a plurality of consecutive time zones t, steps S1 to S10 are repeated for each time zone t. t ) is calculated using the load factor r t-1 The previously calculated load factor r t Also, the probability distribution D t can be defined for each time zone t.

[0065] As described above in detail, according to the prediction method and the simulation device 10 of the present embodiment, the prediction error yt-yh of the first amount of power y supplied from the main power source 2 is calculated. t is the probability distribution D t The prediction error yt-yh t Considering the running cost C t Estimate the running cost C t The expected value E[C t ] is the minimum load factor r t Therefore, the running cost C of the power supply system 1 equipped with the main power supply 2 and the auxiliary power supply 3 can be calculated. t can be reduced.

[0066] [simulation] (Simulation in one time zone t) Simulations were performed using the simulation device 10 under the following three set conditions (Conditions 1 to 3). rated Both are set to 10. Note that the display of units such as power is omitted here. [Table 1] In this simulation, the estimated value y of the first power amount y t If y becomes negative, t = 0, and y t If ≧10, y t = 10. Below, we use the estimated value y t The same processing is performed for

[0067] The following table shows the running cost C calculated by the simulation device 10 according to the present embodiment under the conditions of Table 1. t The expected value E[C(rb t )] and variance V[C(rb t For comparison, the running cost C t The expected value E[C(rh t )] and variance V[C(rh t )] are also shown. [Table 2] The estimated value y of the first power amount y t If it is negative, y t = 0, the probability that the first amount of power y becomes 0 increases under condition 1 where the first amount of power y is small. In this case, the probability of receiving a penalty also increases, and the second cost Cp(y t ) increases. In the simulation using Method II, the initial predicted value yh t The load factor rh of Method I is determined only by tThe results show that running costs are lower when the load factor is slightly lower than the penalty unit cost U penalty This is thought to be due to the low

[0068] Therefore, the rated output P of the main power supply 2 and the auxiliary power supply 3 (SOFC) rated The penalty unit price U penalty The simulation was performed under the following three conditions: [Table 3] In the simulation under conditions 4 to 6, the penalty unit price U penalty When rh is high, the load factor of Method I is t A slightly higher load factor r t The running cost of Method II, C(rb t ) declines. [Table 4]

[0069] (Simulation assuming one day of operation) A simulation was conducted assuming one day of operation. This simulation was conducted with the following settings: (1) The time zone t has 13 frames, from t=1 to 13. (2) In each time zone t, power supply system (SOFC-PV) 1 supplies electricity with a rated output of 10. (3) Initial prediction value yh of the first power amount y t is given by the following function:

number

number

[0070] In Method I, which was performed under these settings, the expected value of the daily running cost E[C(rb t )] is 1112.13, and the variance V[C(rb t )] was 7589.59. The expected value of the daily running cost by Method II, E[C(rh t )] is 1101.46, and the variance V[C(rh t )] was 10084.66. Therefore, Method II reduced the running costs by about 1% compared to Method I. Figure 8(A) shows the running costs C(rh t ) and Fig. 8(B) shows the histogram of the running cost C(rb t ) When comparing the two, the histogram for Method II is slightly lower than the histogram for Method I.

[0071] Figure 9(A) shows the load factor r t The comparison of Method I and Method II is shown in Fig. 9(A). As shown in Fig. 9(A), Method II has a higher load factor r t is getting smaller.

[0072] Figure 9(B) shows the running cost C t The expected value E of the running cost C is compared between Method I and Method II. As shown in FIG. 9B, in each time zone t, Method II has a lower running cost C than Method I. tThe expected value E of is smaller.

[0073] In addition, the first cost Co(r t ), the second cost Cp(y t ), and the third cost Cc(r') are not limited to those defined in the above formula. t ) may be increased in value according to the second power generation amount x(r) of the auxiliary power source 3, and the second cost Cp(y t ) is the predicted power demand f t Total power consumption g t It is sufficient if the third cost Cc(r t ) should increase in accordance with the increase in the load factor of the SOFC. t is simply the first cost Co(r t ), the second cost Cp(y t ), the third cost Cc(r t ) and may be a linear combination of them. t is the first cost Co(r t ), the second cost Cp(y t ), the third cost Cc(r t ) is a value that increases in proportion to the sum of

[0074] In this embodiment, the main power source 2 generates power from solar power. However, this is not limited to this. The main power source 2 may generate power from other renewable energy sources such as wind power or geothermal power, or may generate power using multiple types of renewable energy.

[0075] In this embodiment, the auxiliary power supply 3 is an SOHC. However, this is not limited to this. The auxiliary power supply 3 may be another type of fuel cell. In this case, the load factor r t It is sufficient to calculate the following.

[0076] In addition, the auxiliary power source 3 may be other power generation facilities such as thermal power generation, hydroelectric power generation, and nuclear power generation. In this case, the running cost C(rt ) is the first cost Co(r t ) and the second cost Cp(y), and the third cost Cc(r t ) is not included. In this case, r t can be the amount of fuel instead of the load factor. That is, when the auxiliary power source 3 is a thermal power generation device, for example, the simulation device 10 can calculate the amount of fuel used r t is predicted as a parameter that determines the second amount of power x(r) of the auxiliary power supply 3.

[0077] Furthermore, the hardware configuration and software configuration of the simulation device 10 are merely examples and can be changed or modified as desired.

[0078] The core part of the simulation device 10, which is composed of the CPU 31, main memory 32, external memory 33, operation unit 34, display unit 35, communication unit 36, internal bus 38, etc., may be constructed as a dedicated system as described above, or may be realized by using a normal computer system. For example, the computer program for executing the above-mentioned operations may be stored and distributed on a non-transient computer-readable recording medium (flexible disk, CD-ROM, DVD-ROM, etc.), and the computer program may be installed on a computer to configure the simulation device 10 for executing the above-mentioned processes. Also, the computer program may be stored in a storage device of a server device on a communication network such as the Internet, and the simulation device 10 may be configured by downloading the computer program by a normal computer system.

[0079] When the functions of a computer are realized by sharing between an OS (operating system) and application programs, or by cooperation between the OS and application programs, only the application program portion may be stored in a recording medium or storage device.

[0080] It is also possible to superimpose a computer program on a carrier wave and distribute it via a communication network. For example, the computer program may be posted on a bulletin board system (BBS) on the communication network and distributed via the network. Then, the computer program may be started and executed under the control of the OS in the same way as other application programs, thereby enabling the above-mentioned processing to be performed.

[0081] Various embodiments and modifications of the present invention are possible without departing from the broad spirit and scope of the present invention. The above-described embodiments are for the purpose of explaining the present invention and do not limit the scope of the present invention. That is, the scope of the present invention is indicated by the claims, not the embodiments. Various modifications made within the scope of the claims and the scope of the invention equivalent thereto are considered to be within the scope of the present invention. [Industrial Applicability]

[0082] INDUSTRIAL APPLICABILITY The present invention is applicable to a power supply system that supplies power to consumers using a main power source that generates power from renewable energy and another auxiliary power source. [Explanation of symbols]

[0083] REFERENCE SIGNS LIST 1 power supply system, 2 main power source, 3 auxiliary power source, 4 consumer, 5 adjustment device, 10 simulation device, 11 calculation unit, 12 memory unit, 20 estimated value calculation unit, 21 first cost calculation unit, 22 second cost calculation unit, 23 third cost calculation unit, 24 parameter calculation unit, 31 CPU, 32 main memory, 33 external memory, 34 operation unit, 35 display unit, 36 communication unit, 38 internal bus, 39 program

Claims

1. A prediction method for predicting, by a simulation device, a parameter for determining a first amount of power supplied to a consumer from a main power source that generates power using renewable energy and a second amount of power supplied to the consumer from an auxiliary power source capable of adjusting the power supply, such that a total amount of power is equal to a predicted demand amount of power per unit time of the consumer, generating a plurality of combinations of the estimated value of the first amount of power and the estimated value of the second amount of power based on an initial prediction value of the first amount of power and the predicted amount of power demand by random number calculation assuming that a prediction error of the first amount of power follows a probability distribution; calculating a first cost required for generating the estimated value of the second amount of power by the auxiliary power source for each of the combinations; calculating, for each of the combinations, a plurality of possible values ​​of the first amount of power based on the estimated value of the first amount of power by random number calculation assuming that deviation from the estimated value of the first amount of power follows the probability distribution, and calculating, for each possible value of the first amount of power, a second cost indicating a penalty when the total amount of power is insufficient for the predicted amount of demanded power; calculating the value of the parameter based on the estimated value of the second power amount associated with the combination that minimizes an expected value of the running cost that increases according to the sum of the first cost and the second cost; Forecasting methods.

2. calculating the first cost based on a unit price of fuel used to supply power to the auxiliary power source, a power generation efficiency of the auxiliary power source, and an estimated value of the second amount of power; The prediction method according to claim 1 .

3. When the total amount of power based on a possible value of the first amount of power is insufficient with respect to the predicted amount of power demand, a value proportional to the amount of the shortage is calculated as the second cost; When the total amount of power based on a possible value of the first amount of power is not insufficient for the predicted amount of demanded power, the second cost is set to 0. The prediction method according to claim 1 .

4. the auxiliary power source is a fuel cell, and the parameter is a load factor of the fuel cell; calculating a value of the parameter based on an estimated value of the second power amount related to the combination that minimizes an expected value of a running cost that increases according to a sum of the first cost, the second cost, and a third cost that increases according to an increase in the load factor; The prediction method according to claim 1 .

5. Calculating a load factor of the fuel cell for each time period having a predetermined time span; If the load factor of the time slot of interest has increased compared to the load factor of the time slot immediately preceding the time slot, a value proportional to the amount of increase is calculated as the third cost; When the load factor of the time period of interest has not increased compared to the immediately previous time period, the third cost is set to 0. The prediction method according to claim 4.

6. The main power source is a solar power generation device. The prediction method according to claim 1 .

7. A simulation device that predicts a parameter that determines a first amount of power supplied to a consumer from a main power source that generates power using renewable energy and a second amount of power supplied to the consumer from an auxiliary power source that can adjust the power supply, such that a total amount of power is equal to a predicted amount of power demand per unit time of the consumer, the simulation device comprising: an estimate calculation unit that generates a plurality of combinations of an estimated value of the first amount of power and an estimated value of the second amount of power based on an initial predicted value of the first amount of power and the predicted amount of power demand by random number calculation assuming that a prediction error of the first amount of power follows a probability distribution; a first cost calculation unit that calculates a first cost required for generating the estimated value of the second amount of power by the auxiliary power source for each of the combinations; a second cost calculation unit that calculates, for each of the combinations, a plurality of possible values ​​of the first amount of power based on the estimated value of the first amount of power by random number calculation assuming that a deviation from the estimated value of the first amount of power follows the probability distribution, and calculates, for each possible value of the first amount of power, a second cost that indicates a penalty when the total amount of power is insufficient for the predicted amount of demand power; a parameter calculation unit that calculates a value of the parameter based on an estimated value of the second power amount associated with the combination that minimizes an expected value of a running cost that increases according to a sum of the first cost and the second cost; A simulation device comprising:

Images