Control method, control device, and power system

Abstract

A control method according to the present disclosure comprises a step for executing a first planning method for planning an output of a fuel cell device so as to compensate for a difference between a predicted value of demand power of a power consumer and a predicted value of an output of a photovoltaic generation device. A control device according to the present disclosure comprises: a memory that stores a predicted value of power demand of a power consumer and a predicted value of an output of a photovoltaic generation device; and a planning implement that executes a planning method for planning an output of a fuel cell device so as to compensate for a difference between the predicted value of the demand power of the power consumer and the predicted value of the output of the photovoltaic generation device.

Description

Control method, control device and power system

[0001] The present disclosure relates to a control method, a control device, and a power system.

[0002] Patent Document 1 describes a household distributed power supply system equipped with a solar cell, a fuel cell, and a power storage device. In the household distributed power supply system described in Patent Document 1, a control signal for the output required by the fuel cell is calculated as the difference between the load current and the output current of the solar cell. The power storage device supplies the power shortage caused by a delay in the increase in the fuel cell output.

[0003] Japanese Patent Application Laid-Open No. 2004-208436

[0004] In the prior art, the output of the fuel cell is prone to fluctuation, and the fuel cell is prone to deterioration.

[0005] The present disclosure provides a control method comprising a step of executing a first planning method for planning the output of a fuel cell device so as to compensate for the difference between a predicted value of power demand of a power consumer and a predicted value of output of a solar power generation device.

[0006] According to the present disclosure, deterioration of the fuel cell device can be suppressed.

[0007] Fig. 1 is a configuration diagram of a power system according to an embodiment of the present disclosure. Fig. 2 is a functional block diagram of the control device and database shown in Fig. 1. Fig. 3 is a flowchart showing a process for switching between a first planning method and a second planning method. Fig. 4A is a graph showing fluctuations in various powers and energy storage rates in the power system of this embodiment when the first planning method is executed. Fig. 4B is a graph showing fluctuations in various powers and energy storage rates in the power system of this embodiment when the second planning method is executed.

[0008] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. The present disclosure is not limited to the following embodiments.

[0009] 1 is a configuration diagram of a power system according to an embodiment of the present disclosure. The power system 200 includes a control device 10, a database 20, a solar power generation system 30, a fuel cell system 40, and a power storage system 50. The solar power generation system 30, the fuel cell system 40, and the power storage system 50 are connected to a load 300 via a power line 102. Electric power generated by the solar power generation system 30 and the fuel cell system 40 is supplied to the load 300 via the power line 102. As needed, the electric power generated by the solar power generation system 30 and the fuel cell system 40 is stored in the power storage system 50, and the electric power is supplied from the power storage system 50 to the load 300.

[0010] A commercial power source 100 is connected to the power line 102. The power system 200 may be interconnected with the commercial power source 100. Power may be purchased from the commercial power source 100, or surplus power may be sold to the commercial power source 100 (reverse power flow). However, according to the power system 200 of this embodiment, power shortages are unlikely to occur, and surplus power is also unlikely to occur.

[0011] The solar power generation apparatus 30 includes a solar cell module 31, a first PCS (Power Conditioning Subsystem) 32, a first power meter 33, and a first control device 34. The solar power generation apparatus 30 may include a plurality of solar cell modules 31. The solar cell module 31 may be configured with a plurality of solar cell panels. The first PCS 32 converts the power generated by the solar cell module 31 from DC power to AC power and outputs the AC power to the power line 102. The first power meter 33 detects the power output from the first PCS 32, i.e., the power generated by the solar power generation apparatus 30. The first control device 34 is communicatively connected to the control device 10. The first control device 34 controls the solar cell module 31 or the first PCS 32 in accordance with instructions received from the control device 10. The first control device 34 may be a computer including a storage device in which a program required for controlling the solar power generation apparatus 30 is stored and a processor that reads and executes the program from the storage device.

[0012] The fuel cell system 40 includes a fuel cell unit 41, a second PCS 42, a second power meter 43, and a second control device 44. The fuel cell system 40 may include multiple fuel cell units 41 that can operate independently of each other. The power generated by the fuel cell system 40 can be changed by adjusting the number of operating fuel cell units 41 and / or their output. The second PCS 42 converts the power generated by the fuel cell unit 41 from DC power to AC power and outputs it to the power line 102. The second power meter 43 detects the power output from the second PCS 42, i.e., the power generated by the fuel cell system 40. The second control device 44 is communicatively connected to the control device 10. The second control device 44 controls the fuel cell unit 41 or the second PCS 42 in accordance with instructions received from the control device 10. The second control device 44 may be a computer including a storage device that stores a program required for controlling the fuel cell system 40 and a processor that reads and executes the program from the storage device.

[0013] The fuel cell in the fuel cell device 40 may be a polymer electrolyte fuel cell, a solid oxide fuel cell, a phosphoric acid fuel cell, a molten carbonate fuel cell, or the like.

[0014] The power storage device 50 includes a storage battery 51, a third PCS 52, a third wattmeter 53, and a third control device 54. The storage battery 51 may be a secondary battery. The power storage device 50 may include a capacitor instead of the storage battery 51. The third PCS 52 is a bidirectional PCS. When power is stored in the storage battery 51, the third PCS 52 converts AC power to DC power. When power is discharged from the storage battery 51, the third PCS 52 converts DC power to AC power. The third wattmeter 53 detects the power output from the third PCS 52 and the power input to the third PCS 52. In other words, the third wattmeter 53 detects the charging power and discharging power of the power storage device 50. The third control device 54 is communicatively connected to the control device 10. The third control device 54 controls the storage battery 51 or the third PCS 52 according to instructions received from the control device 10. The third control device 54 may be a computer including a storage device that stores a program required for controlling the power storage device 50, and a processor that reads and executes the program from the storage device.

[0015] The third control device 54 sequentially transmits the state of charge (SOC) of the power storage device 50 to the control device 10. The state of charge (SOC) of the power storage device 50 is stored in the database 20 in chronological order.

[0016] The load 300 represents a power consumer such as a factory. The fourth power meter 302 detects, for example, the power consumption of the factory. The fourth power meter 302 may be part of the power system 200.

[0017] The control device 10 issues control instructions to each of the first control device 34, the second control device 44, and the third control device 54 to control the operation of the photovoltaic power generation device 30, the fuel cell device 40, and the power storage device 50. Specifically, the control device 10 transmits information necessary for the respective control operations to each of the first control device 34, the second control device 44, and the third control device 54. The control device 10 is a computer equipped with a storage device that stores programs necessary for controlling these devices and a processor that reads and executes the programs from the storage device. The control device 10 acquires detection values ​​from the first wattmeter 33, the second wattmeter 43, the third wattmeter 53, and the fourth wattmeter 302 at a predetermined sampling period (e.g., every 30 seconds) and transmits them to the database 20. From these detection values, the power generated by the photovoltaic power generation device 30, the power generated by the fuel cell device 40, the charging power of the power storage device 50, the discharging power of the power storage device 50, and the power consumption of the load 300 can be determined.

[0018] The database 20 is configured by a server computer and is communicably connected to the control device 10. The database 20 stores data in chronological order, such as the power generated by the solar power generation device 30, the power generated by the fuel cell device 40, the charging power of the power storage device 50, the discharging power of the power storage device 50, the power consumption of the load 300, and the power storage rate of the power storage device 50. The amount of power (kWh) can be calculated from the time-series data.

[0019] The control device 10 may be a cloud computer or an edge computer. The database 20 may be a database in a cloud server or an edge server. The cloud computer and the cloud server may be located at a location away from the locations where the photovoltaic power generation device 30, the fuel cell device 40, and the power storage device 50 are installed. The cloud computer as the control device 10 is communicatively connected to the first control device 34, the second control device 44, and the third control device 54 via a communication network such as the Internet. The edge computer and the edge server may be located near the locations where the photovoltaic power generation device 30, the fuel cell device 40, and the power storage device 50 are installed.

[0020] 2 is a functional block diagram of the control device 10 and database 20 shown in FIG. 1. The database 20 stores factory operation data, weather data, power data, and power storage data. The control device 10 creates a prediction model through machine learning using the various data stored in the database 20, and uses the prediction model to create operation instructions for the fuel cell device 40. The prediction model includes a prediction model for predicting the power demand of the factory and a prediction model for predicting the power generated by the solar power generation device 30.

[0021] In the database 20, the factory operation data is data entered by the administrator from the terminal 21. The factory operation data is a combination of data indicating whether the factory is operating or not and date data. The data indicating whether the factory is operating or not is data indicating a non-business day or an operating day. The factory operation data makes it possible to identify whether a specific day in the past was a non-business day or an operating day. There is a correlation between the power consumption of a factory and whether the factory is operating or not. The power consumption of a factory is high on operating days, but low on non-business days. Therefore, the factory operation data can be used as learning data when creating a prediction model for predicting the power demand of a factory.

[0022] Even when the load 300 (FIG. 1) is a facility other than a factory, there is a correlation between power consumption and whether or not it is operating. Examples of facilities other than factories include commercial facilities, hospitals, schools, and office buildings. For example, when the load 300 is a commercial facility such as a department store, the operation data can be a combination of date data and data indicating one of "holiday," "business day and weekday," and "business day and holiday." A commercial facility consumes more power on business days, but less power on holiday days. Even on the same business day, there is a difference between power consumption on weekdays and power consumption on holidays.

[0023] In the database 20, weather data is input from a weather information service 22 via a communication network such as the Internet. The weather data may be input by an administrator from the terminal 21. The weather data is a combination of data indicating the nature of the weather and date data. The data indicating the nature of the weather is, for example, data combining a time period and the intensity of sunlight. The weather data allows the intensity of sunlight for each time period in the past to be determined. The intensity of sunlight is expressed in multiple intensities, for example, from an intensity corresponding to clear skies to an intensity corresponding to rain. There is a correlation between the intensity of sunlight and the power generated by the solar power generation device 30. When there is sunlight, the power generation power (kW) is high, and when there is no sunlight, the power generation power is low. Therefore, the weather data indicating the intensity of sunlight can be used as learning data when creating a prediction model for predicting the power generation power of the solar power generation device 30. Instead of or in addition to the intensity of sunlight, the weather data may include the duration of sunlight for each time period. There is a correlation between the duration of sunlight and the power generation amount (kWh) of the solar power generation device 30.

[0024] Data indicating weather conditions may include data on a combination of time period and temperature. For example, when the temperature is high, the power consumption of air conditioners increases, and so does the power consumption of factories. Even when the temperature is low, the power consumption of air conditioners increases, and so does the power consumption of factories. Therefore, weather data indicating temperatures can be used as learning data when creating a prediction model for predicting the power demand of factories.

[0025] The data indicating the weather conditions is not limited to the above, and may include other data that is correlated with the power generated by the solar power generation device 30, or other data that is correlated with the power demand.

[0026] In the database 20, the power data is data input from the power meters 33, 43, 53, and 302 via the control device 10. The power data may be time-series data for the power generated by the photovoltaic power generation system 30, the power generated by the fuel cell system 40, the charging power of the power storage system 50, the discharging power of the power storage system 50, and the power consumption of the load 300. The power data makes it possible to identify the power generated by the photovoltaic power generation system 30, the power generated by the fuel cell system 40, the charging power of the power storage system 50, the discharging power of the power storage system 50, and the power consumption of the load 300 at a specific time in the past. The power data indicating the power consumption of the load 300 can be used as learning data when creating a prediction model for predicting the power demand of the factory. The data on the power generated by the photovoltaic power generation system 30 can be used as learning data when creating a prediction model for predicting the power generation of the photovoltaic power generation system 30.

[0027] In the database 20, the power storage data is data input from the power storage device 50 via the control device 10. The power storage data is data indicating the power storage rate of the power storage device 50.

[0028] The control device 10 includes a prediction unit 12 , a planning unit 14 , and a control unit 16 .

[0029] The prediction unit 12 has a power demand prediction unit 12a and a photovoltaic power generation prediction unit 12b. In the power demand prediction unit 12a, the predicted value of power demand is predicted based on the correlation between actual weather conditions and actual values ​​of power demand, and on the weather forecast. In the power demand prediction unit 12a, the predicted value of power demand is predicted based on the correlation between actual weather conditions and actual values ​​of power demand of the factory, the weather forecast, and the factory operation plan. In the photovoltaic power generation prediction unit 12b, the predicted value of output of the photovoltaic power generation device 30 is predicted based on the correlation between actual weather conditions and actual values ​​of output of the photovoltaic power generation device 30, and on the weather forecast.

[0030] The power demand prediction unit 12a may be a program module including a prediction model for predicting the power demand of the load 300. The prediction model in the power demand prediction unit 12a is created using actual power demand records, weather records, and factory operation records. Power data indicating the power consumption of the load 300 is used as the actual power demand records. Weather data is used as the weather records. The weather records are weather records for areas where power consumers such as factories are located. Factory operation data is used as the factory operation records. The prediction model is, for example, a regression model. The power data, weather data, and factory operation data are supervised data in machine learning.

[0031] A predicted value of the power demand (power consumption) of the factory can be obtained by inputting a weather forecast and a factory operation plan into the prediction model in the power demand prediction unit 12a. That is, a predicted value of the power demand of the factory, which is a dependent variable, is obtained using the weather forecast and the factory operation plan as explanatory variables. For example, data on the temperature at the time or time period when the predicted value of the power demand should be obtained is used as the weather forecast. The weather forecast may include data indicating the intensity of sunlight or the duration of sunlight. The weather forecast is a weather forecast for an area where power consumers such as factories are located. Data indicating whether the factory will be operating on the date and time when the predicted value of the power demand should be obtained is used as the factory operation plan.

[0032] The solar power generation prediction unit 12b may be a program module including a prediction model for predicting the power generation of the solar power generation device 30. The prediction model in the solar power generation prediction unit 12b is created using the power generation record and weather record of the solar power generation device 30. Power data indicating the power generation of the solar power generation device 30 is used as the power generation record of the solar power generation device 30. Weather data is used as the weather record. The prediction model is, for example, a regression model. The power data and weather data are supervised data in machine learning.

[0033] By inputting the weather forecast into the prediction model in the solar power generation prediction unit 12b, a predicted value of the output (power generation) of the solar power generation device 30 can be obtained. That is, the predicted value of the output of the solar power generation device 30, which is the objective variable, is obtained using the weather forecast as an explanatory variable. As the weather forecast, for example, data indicating the intensity or duration of sunlight at the time or time zone when the predicted value of the output of the solar power generation device 30 should be obtained is used.

[0034] To create each of the above-described prediction models in the prediction unit 12, various data from a predetermined period (e.g., three months to one year) is used. The prediction models may be updated daily, which is expected to improve prediction accuracy. The prediction models may be stored in the database 20.

[0035] The process of obtaining predicted values ​​by the prediction unit 12 is executed at a predetermined cycle. For example, a plurality of predicted values ​​from the present to 24 hours later are obtained at a cycle of once every hour. The plurality of predicted values ​​are, for example, 24 predicted values ​​at hourly intervals.

[0036] A process of correcting the predicted value may be performed. For example, the predicted value of the power demand may be corrected in the following manner. That is, the predicted value obtained from the prediction model may be corrected based on the actual value of the power demand for the most recent first period and the predicted value of the power demand for the second period corresponding to the first period. The most recent first period is a period from the current time back a predetermined time (30 minutes to 1 hour). For example, if the current time is 10:00, the most recent first period is the period from 9:00 to 10:00. The actual value of the power demand for this period is stored in chronological order in the database 20 as the power consumption of the load 300. The actual value of the power demand for the first period may be an average value for the first period. Next, the second period corresponding to the first period is a period corresponding to the first period in the 24-hour period for which the 24 predicted values ​​were obtained. For example, when the first period is from 9:00 to 10:00, the corresponding second period is from 9:00 to 10:00.

[0037] For example, assume that the actual power demand value for the first period from 9:00 to 10:00 is 100 kW. Meanwhile, assume that the predicted power demand value for the second period from 9:00 to 10:00 is 90 kW. In this case, the actual value exceeds the predicted value by 10 kW. Therefore, a correction is performed by adding 10 kW to the predicted power demand value for 10:00 to 11:00. This improves the prediction accuracy. The difference between the actual value and the predicted value may be used directly for the correction, or a predetermined percentage (e.g., 80%) of the difference between the actual value and the predicted value may be used for the correction. The degree of correction is arbitrary. Furthermore, of the predicted values ​​for 24 future points, only the most recently used predicted value may be corrected, or multiple predicted values ​​including the most recently used predicted value may be corrected.

[0038] The predicted value of the power generation power of the photovoltaic power generation device 30 may be corrected using the same method. That is, the predicted value obtained from the prediction model may be corrected based on the actual value of the output of the photovoltaic power generation device 30 for the most recent third period and the predicted value of the output of the photovoltaic power generation device 30 for a fourth period corresponding to the third period. This can improve the prediction accuracy. The third period may be the same period as the first period or may be a different period. The fourth period may be the same period as the second period or may be a different period.

[0039] The predicted values ​​calculated by the prediction unit 12 are stored in the database 20. With this configuration, it is possible to execute the prediction process for calculating the predicted values ​​and the planning process for calculating the planned values ​​asynchronously. The server that serves as the database 20 is an example of a storage device that stores the predicted values ​​of power demand and the predicted values ​​of the output of the photovoltaic power generation device 30.

[0040] The planner 14 includes a fuel cell output planner 14a. The fuel cell output planner 14a may be a program module for calculating a planned value of the output (power generation) of the fuel cell device 40. The fuel cell output planner 14a acquires a predicted value Pd of power demand and a predicted value Ppv of the output of the photovoltaic power generation device 30 from the predictor 12, and calculates a planned value of the output of the fuel cell device 40 using these predicted values. For example, when the current time is 8:50, a predicted value of power demand and a predicted value of the output of the photovoltaic power generation device 30 for the time period from 9:00 to 10:00 are created in the predictor 12, and a planned value of the output of the fuel cell device 40 is calculated from these predicted values. The predicted value and the planned value may be expressed in terms of instantaneous power (kW) or the amount of power per unit time (kWh).

[0041] In this embodiment, the planner 14 plans the output of the fuel cell device 40 so as to compensate for the difference between the predicted value Pd of power demand and the predicted value Ppv of the output of the photovoltaic power generation device 30. By planning the output of the fuel cell device 40 using the predicted value, fluctuations in the output of the fuel cell device 40 can be suppressed compared to when the output of the fuel cell device 40 is planned based only on the most recent actual values. Since there is a correlation between the lifespan of the stack of the fuel cell device 40 and the frequency of fluctuations in the output of the fuel cell device 40, suppressing fluctuations in the output of the fuel cell device 40 extends the lifespan of the fuel cell device 40.

[0042] The planned value Pfc of the output of the fuel cell device 40 is calculated, for example, as the difference (Pd-Ppv) between the predicted value Pd of power demand and the predicted value Ppv of the output of the photovoltaic power generation device 30. Therefore, when the current time is 8:50, the planned value Pfc of the output of the fuel cell device 40 for the time slot from 9:00 to 10:00 can be (Pd-Ppv). If the difference is zero or less, that is, if (Pd-Ppv)≦0, the planned value Pfc of the output of the fuel cell device 40 is zero.

[0043] According to the method of this embodiment, a predicted value Pd of power demand and a predicted value Ppv of the output of the photovoltaic power generation device 30 are calculated. Therefore, even if a method of determining the output of the fuel cell device 40 based on the most recent data would cause the output of the fuel cell device 40 to be varied, the method of this embodiment may determine that it is better not to vary the output of the fuel cell device 40. This difference is due to the poor tracking ability of the output of the fuel cell device 40.

[0044] The process of calculating the planned values ​​by the planning unit 14 is executed at a predetermined interval. For example, the planned values ​​required for operation in the immediate time period are calculated once every hour.

[0045] The planner 14 is an example of a planner that executes a planning method for planning the output of the fuel cell device 40 so as to compensate for the difference between the predicted value Pd of power demand and the predicted value Ppv of the output of the photovoltaic power generation device 30 .

[0046] In this specification, "compensating for the difference" means both compensating for the entire difference and compensating for only a portion of the difference. In the above example, the difference between the predicted value Pd of power demand and the predicted value Ppv of the output of the photovoltaic power generation device 30 is regarded as the planned value Pfc of the output of the fuel cell device 40, and therefore the entire difference is compensated for by the output of the fuel cell device 40.

[0047] The control unit 16 includes a fuel cell correction control unit 16a, a power storage control unit 16b, and a solar cell output suppression unit 16c.

[0048] In the control unit 16, the fuel cell correction control unit 16a may be a program module that determines whether or not the planned value Pfc of the output of the fuel cell device 40 needs to be corrected, and corrects the planned value Pfc as necessary. The fuel cell correction control unit 16a acquires the planned value Pfc of the output of the fuel cell device 40 from the planner 14, and also acquires the actual value Rd of the power demand, the actual value Rpv of the output of the photovoltaic power generation device 30, and the state of charge (SOC) of the power storage device 50 from the database 20. The actual value Rd of the power demand is the power consumption of the load 300. Specifically, the average value of the power consumption over a predetermined period can be used as the actual value Rd of the power demand. The average value of the power generated by the photovoltaic power generation device 30 over the predetermined period can be used as the actual value Rpv of the output of the photovoltaic power generation device 30. The average value is, for example, a moving average value. The predetermined period is, for example, the period from 30 minutes ago to the time when the most recent data was acquired (approximately the current time). The charge rate of the power storage device 50 is the charge rate of the power storage device 50 at the time when the most recent data was acquired (approximately the current time).

[0049] It is desirable for the power system 200 to avoid purchasing power from the commercial power source 100 and selling power to the commercial power source 100 (reverse power flow) as much as possible, and it is also desirable for the power storage device 50 to be able to charge and discharge at all times. If the power storage device 50 is in a fully charged state or a fully discharged state, it becomes difficult for the power storage device 50 to balance supply and demand in the power system 200. In this embodiment, the fuel cell correction control unit 16a corrects the planned value of the output of the fuel cell device 40, taking into account the power storage rate of the power storage device 50, so as to prevent the power storage rate of the power storage device 50 from becoming a fully charged state or a fully discharged state.

[0050] The fuel cell correction control unit 16a corrects the planned value Pfc of the output of the fuel cell device 40 based on the storage rate (SOC) of the power storage device 50 predicted from the difference between the actual value Rd of power demand and the sum of the actual value Rpv of the output of the photovoltaic power generation device 30 and the planned value Pfc of the output of the fuel cell device 40. This makes it possible to create a command value for the output of the fuel cell device 40 while taking into account the storage rate of the power storage device 50, and ultimately to maintain the power storage device 50 in a state where it can be charged and discharged.

[0051] Specifically, the value (Rd-Rpv-Pfc) is calculated by subtracting the sum of the actual value Rpv of the output of the photovoltaic power generation device 30 and the planned value Pfc of the output of the fuel cell device 40 from the actual value Rd of the power demand. This value may be zero, a negative value, or a positive value.

[0052] When the above value (Rd - Rpv - Pfc) is a negative value, prediction based on the most recent actual value indicates a high possibility of excess power in the future. Therefore, the storage rate after a predetermined time (e.g., one hour) when the power storage device 50 is charged with power corresponding to the absolute value of the value (Rd - Rpv - Pfc) is predicted from the most recent storage rate. The storage rate after the predetermined time is calculated from the most recent storage rate of the power storage device 50, the capacity of the power storage device 50, and the charging power of the power storage device 50. If the storage rate after the predetermined time exceeds a predetermined upper limit value that is lower than the fully charged state (100%), the planned output value Pfc of the fuel cell device 40 is corrected so that the storage rate after the predetermined time is equal to or lower than the upper limit value. For example, the planned output value Pfc of the fuel cell device 40 is corrected by subtracting a correction value from the planned output value Pfc of the fuel cell device 40 or multiplying it by a correction value less than 1. As a result, the fuel cell device 40 is operated at an output lower than the planned value Pfc. The upper limit of the charge rate is set to, for example, 80% of the capacity of the power storage device 50 .

[0053] If the above value (Rd - Rpv - Pfc) is a positive value, it can be said that, based on the most recent actual value, there is a high possibility of a power shortage in the future. Therefore, the storage rate after a predetermined time (e.g., one hour) when power equivalent to the value (Rd - Rpv - Pfc) is discharged from the power storage device 50 is predicted from the most recent storage rate. If the storage rate after the predetermined time falls below a predetermined lower limit value that is higher than a fully discharged state (0%), the planned output value Pfc of the fuel cell device 40 is corrected so that the storage rate after the predetermined time is equal to or higher than the lower limit value. For example, the planned output value Pfc of the fuel cell device 40 is corrected by adding a correction value to the planned output value Pfc of the fuel cell device 40 or multiplying it by a correction value greater than 1. As a result, the fuel cell device 40 will operate at an output greater than the planned value Pfc. The lower limit of the storage rate is set to, for example, 20% of the capacity of the power storage device 50.

[0054] If the above value (Rd - Rpv - Pfc) is zero, the planned value Pfc is not corrected. Also, if the charge rate after a predetermined time is equal to or greater than the lower limit and equal to or less than the upper limit, the planned value Pfc is not corrected.

[0055] The fuel cell correction control unit 16a uses the planned value Pfc or the corrected planned value Pfc' to create a command value for the output of the fuel cell device 40. The command value is sent to the second control device 44 of the fuel cell device 40. The second control device 44 controls the operation of the fuel cell unit 41 so that power is generated in accordance with the command value.

[0056] The process of correcting the planned value Pfc of the output of the fuel cell device 40 may be executed in synchronization with the process of creating the planned value Pfc, or may be executed in a control cycle shorter than that of the process of creating the planned value Pfc. The planned value Pfc may be corrected when it is determined that the power storage rate of the power storage device 50 will exceed the upper limit value or fall below the lower limit value.

[0057] In the control unit 16, the power storage control unit 16b may be a program module for controlling the power storage device 50 to adjust the charging and discharging power of the power storage device 50. The power storage control unit 16b acquires the actual value Rd of power demand, the actual value Rpv of the output of the photovoltaic power generation device 30, and the actual value Rfc of the output of the fuel cell device 40 from the database 20. Since the output of the fuel cell device 40 follows the command value, the command value for the output of the fuel cell device 40 may be considered to be the actual value Rfc of the output of the fuel cell device 40.

[0058] The power storage control unit 16b controls the power storage device 50 so as to compensate for the difference between the actual value Rd of power demand and the sum of the actual value Rpv of the output of the photovoltaic power generation device 30 and the actual value Rfc of the output of the fuel cell device 40 by charging or discharging the power storage device 50. With this configuration, it is possible to achieve a balance between supply and demand in the power system 200, and therefore it is possible to avoid purchasing power from or selling power to the commercial power source 100 as much as possible. This is advantageous from the perspective of avoiding the use of power from the commercial power source 100 and increasing the renewable energy utilization rate.

[0059] Specifically, the actual value Rd of the power demand is subtracted by the actual value Rpv of the output of the photovoltaic power generation device 30 and the actual value Rfc of the output of the fuel cell device 40 to calculate a value (Rd - Rpv - Rfc). This value is zero, a negative value, or a positive value. When the value (Rd - Rpv - Rfc) is negative, there is surplus power, and a command value for charging power of the power storage device 50 is created so that the power storage device 50 is charged with the surplus power. When the value (Rd - Rpv - Rfc) is positive, there is a power shortage, and a command value for discharging power of the power storage device 50 is created so that the power shortage is compensated for by discharging the power storage device 50. When the value (Rd - Rpv - Rfc) is zero, a command value for charging and discharging power of the power storage device 50 is created so that the charging and discharging power is zero. The actual value Rd of the power demand is the power consumption of the load 300. The most recent power consumption can be used as the actual value Rd of the power demand. The most recent output can be used as the actual value Rpv of the output of the photovoltaic power generation device 30 and the actual value Rfc of the output of the fuel cell device 40. The most recent power consumption and the most recent output are values ​​included in the power data stored in the database 20.

[0060] A command value for the charge / discharge power of the power storage device 50 is sent from the power storage control unit 16b to the third control device 54 of the power storage device 50. The third control device 54 controls the operation of the power storage device 50 so that the power storage device 50 is charged with power according to the command value or discharged with power according to the command value. The charge / discharge power of the power storage device 50 is adjusted to a desired value by controlling the third PCS 52.

[0061] In the control unit 16, the solar cell output suppression unit 16c may be a program module for suppressing the output of the solar power generation device 30. When the power storage rate of the power storage device 50 exceeds a first threshold, the solar cell output suppression unit 16c generates a command value for the solar power generation device 30 to suppress the output of the solar power generation device 30. When the power storage rate of the power storage device 50 falls below a second threshold, the suppression of the output of the solar power generation device 30 is released. The first threshold for the power storage rate is, for example, 90% of the capacity of the power storage device 50. The second threshold for the power storage rate is, for example, 85% of the capacity of the power storage device 50. With this configuration, it is possible to prevent the power storage device 50 from reaching a fully charged state, and therefore it is possible to avoid selling power to the commercial power source 100 (reverse power flow) as much as possible.

[0062] The process of suppressing the output of the solar power generation device 30 is as follows: The solar cell output suppression unit 16c sets a value equal to or less than the difference between the sum of the actual value Rd of power demand and the chargeable power Sc of the power storage device 50 and the actual value Rfc of the output of the fuel cell device 40 as the upper limit of the output of the solar power generation device 30. With this configuration, it becomes possible to use the power generated by the solar power generation device 30 without any waste.

[0063] Specifically, a value (Rd+Sc-Rfc) is calculated by subtracting the actual output value Rfc (kW) of the fuel cell device 40 from the sum of the actual power demand value Rd (kW) and the chargeable power Sc (kW) of the power storage device 50. The value (Rd+Sc-Rfc) or a value less than this is set as the upper limit of the output of the solar power generation device 30. The actual power demand value Rd is, for example, power consumption at an arbitrary point in time in the past, and may be the average value of power demand in the most recent time period (e.g., the most recent hour), or the minimum value of power demand in the most recent time period. The chargeable power Sc of the power storage device 50 is any power equal to or less than the maximum charging power of the power storage device 50. The maximum charging power of the power storage device 50 is a design value for the power storage device 50 determined depending on the type of storage battery 51, etc. In other words, the power storage device 50 is not charged with power exceeding the maximum charging power. The actual output value Rfc of the fuel cell device 40 is, for example, the most recent output value of the fuel cell device 40 .

[0064] When the output of the solar power generation device 30 is not suppressed, the first PCS 32 performs MPPT (Maximum Power Point Tracking) control so that the maximum output is supplied from the solar power generation device 30 to the power line 102. On the other hand, when an upper limit value for the output of the solar power generation device 30 is set, the first PCS 32 operates the solar power generation device 30 at an operating point that is different from the optimal operating point on the IV curve so that the output of the solar power generation device 30 does not exceed the upper limit value. In other words, the first PCS 32 does not perform MPPT control. In this embodiment, the first PCS 32 has such a function.

[0065] As described above, in this embodiment, a prediction model for predicting the power demand of a factory and a prediction model for predicting the power generated by the photovoltaic power generation device 30 are used. Creating a prediction model requires a sufficient number of days of various data. Until a prediction model with sufficient accuracy is created, a planned value for the output of the fuel cell device 40 can be calculated using the most recent data on power demand and the most recent data on the output of the photovoltaic power generation device 30. Specifically, the difference between the most recent power demand and the most recent output of the photovoltaic power generation device 30 is calculated as the planned value for the output of the fuel cell device 40. The most recent power demand is, for example, the average value of the power demand over the most recent hour. The most recent output of the photovoltaic power generation device 30 is, for example, the average value of the output of the photovoltaic power generation device 30 over the most recent hour. The average value is, for example, a moving average value. The most recent power demand may be the minimum value of the power demand in the most recent time period (e.g., the most recent hour).

[0066] In this specification, a method of calculating a planned value for the output of the fuel cell device 40 using a prediction model is defined as a first planning method, and a method of calculating a planned value for the output of the fuel cell device 40 using only the most recent data is defined as a second planning method.

[0067] FIG. 3 is a flowchart showing a process for switching between the first planning method and the second planning method. Each process shown in FIG. 3 is executed by the control device 10. In step S1, the second planning method is executed. In step S2, it is determined whether a predetermined period has elapsed. After the predetermined period has elapsed, in step S3, switching from the second planning method to the first planning method is performed. In step S4, the first planning method is executed. That is, after executing the second planning method for a predetermined period, switching from the second planning method to the first planning method is performed. The predetermined period is, for example, three months to one year.

[0068] Prior to the first planning method, a second planning method is executed to plan the output of the fuel cell system 40 so as to compensate for the difference between the actual value Rd of the factory's power demand and the actual value Rpv of the output of the photovoltaic power generation system 30. This makes it possible to collect learning data necessary for creating a prediction model. This improves the accuracy of predictions when the first planning method is executed. After the second planning method is executed for a predetermined period, the method for planning the output of the fuel cell system 40 switches from the second planning method to the first planning method. In the first planning method, the predicted value Pd of the factory's power demand is predicted based on the actual value Rd of the factory's power demand. In the first planning method, the predicted value Ppv of the output of the photovoltaic power generation system 30 is predicted based on the actual value Rpv of the output of the photovoltaic power generation system 30.

[0069] According to the flowchart of Fig. 3, the switch from the second planning method to the first planning method is performed automatically after a predetermined period of time has elapsed, but the switch from the second planning method to the first planning method may also be performed in response to an input from an administrator.

[0070] FIG. 4A is a graph showing fluctuations in various power levels and energy storage rates in the power system of this embodiment when the first planning method is executed. FIG. 4B is a graph showing fluctuations in various power levels and energy storage rates in the power system of this embodiment when the second planning method is executed. The horizontal axis of the graph represents the passage of time. The vertical axis on the left side of the upper graph represents power (kW). The vertical axis on the right side of the upper graph represents energy storage rate (%). The vertical axis of the lower graph represents purchased power from a commercial power source or reverse flow power (kW) to the commercial power source. The power consumption W1 of the load 300 and the output of the photovoltaic power generation device 30 are data obtained from an actual facility. The output W3 of the fuel cell device 40, the charge / discharge power W4 of the energy storage device 50, the energy storage rate W5, and the reverse flow power W6 are data obtained by computer simulation.

[0071] The power consumption W1 of the load 300 is consistent between Figures 4A and 4B. In the examples of Figures 4A and 4B, no upper limit is set for the output W2 of the solar power generation device 30, so the output W2 of the solar power generation device 30 is also consistent between Figures 4A and 4B.

[0072] As shown in Figure 4B, when the output of the fuel cell device 40 was planned based on the most recent data without prediction using a prediction model, the output W3 of the fuel cell device 40 fluctuated significantly over time. The storage rate W5 of the power storage device 50 was between 80% and 100% for most of the time period. Therefore, during the period when the output W2 of the photovoltaic power generation device 30 exceeded the power consumption W1 (particularly from June 13 to June 14), the charging power W4 of the power storage device 50 could not be increased. In the example of Figure 4B, reverse flow power W6 occurred.

[0073] In contrast, as shown in FIG. 4A , when predictions were made using the prediction model, fluctuations in the output W3 of the fuel cell device 40 over time were small. For example, from June 11 to June 12, the output W3 was maintained at approximately 100 kW. Because the fluctuations in the output of the fuel cell device 40 were small, damage to the stack was suppressed. This leads to a longer lifespan of the fuel cell device 40. The energy storage rate W5 of the power storage device 50 was maintained between 20% and 80% throughout the entire period. Therefore, during the period when the output W2 of the solar power generation device 30 exceeded the power consumption W1 (particularly from June 13 to June 14), the charging power W4 of the power storage device 50 was increased, and the power storage device 50 absorbed the surplus power. In the example of FIG. 4A , no reverse power flow occurred.

[0074] (Other Embodiments) (Additional Notes) The above description of the embodiments discloses the following techniques.

[0075] (Technology 1) A control method including a step of executing a first planning method for planning the output of a fuel cell device so as to compensate for the difference between a predicted value of power demand of an electric power consumer and a predicted value of output of a photovoltaic power generation device. With this configuration, it is possible to suppress deterioration of the fuel cell device.

[0076] (Technology 2) The control method according to Technology 1, wherein the planned value of the output of the fuel cell device is corrected based on the storage rate of the power storage device predicted from the difference between the actual value of the power demand of the power consumer and the sum of the actual value of the output of the solar power generation device and the planned value of the output of the fuel cell device. With this configuration, it is possible to create a command value for the output of the fuel cell device while taking into account the storage rate of the power storage device, and ultimately to maintain the power storage device in a state where it can be charged and discharged.

[0077] (Technology 3) The control method according to Technology 1 or 2, wherein an upper limit value of the output of the solar power generation device is set to a value equal to or less than the difference between the sum of the actual value of the power demand of the power consumer and the chargeable power of the power storage device that is equal to or less than the maximum charge power of the power storage device, and the actual value of the output of the fuel cell device. With this configuration, it becomes possible to use the power generated by the solar power generation device without waste.

[0078] (Technology 4) The control method according to Technology 1, wherein the power storage device is controlled so as to compensate for a difference between an actual value of the power demand of the power consumer and the sum of an actual value of the output of the photovoltaic power generation device and an actual value of the output of the fuel cell device by charging or discharging the power storage device. With this configuration, it is possible to balance supply and demand in the power system, and therefore it is possible to avoid purchasing power from and selling power to commercial power sources as much as possible.

[0079] (Technology 5) A control method according to Technology 1, wherein the predicted value of the power demand of the power consumer is a value predicted based on a correlation between actual weather conditions in the area where the power consumer is located and the actual value of the power demand of the power consumer, and a weather forecast for the area where the power consumer is located, and is corrected based on the actual value of the power demand of the power consumer for a most recent first period and the predicted value of the power demand of the power consumer for a second period corresponding to the first period.

[0080] (Technology 6) The control method according to Technology 5, wherein the power consumer includes a factory, and the predicted value of the power demand of the power consumer is a value predicted based on a correlation between the operation record of the factory, the weather record, and the actual value of the power demand of the power consumer, the weather forecast, and the operation plan of the factory, and corrected based on the actual value of the power demand of the power consumer in the first period and the predicted value of the power demand of the power consumer in the second period.

[0081] (Technology 7) The control method according to Technology 1, wherein the predicted value of the output of the solar power generation device is a value predicted based on a correlation between actual weather conditions in the area where the power consumer is located and the actual value of the output of the solar power generation device, and a weather forecast for the area where the power consumer is located, and is corrected based on the actual value of the output of the solar power generation device for a most recent third period and the predicted value of the output of the solar power generation device for a fourth period corresponding to the third period.

[0082] (Technology 8) The control method according to Technology 1 further includes the steps of: executing, before the first planning method, a second planning method for planning the output of the fuel cell device so as to compensate for a difference between an actual value of the power demand of the power consumer and an actual value of the output of the photovoltaic power generation device; and executing the second planning method for a predetermined period and then switching to the first planning method, wherein the predicted value of the power demand of the power consumer is predicted based on the actual value of the power demand of the power consumer, and the predicted value of the output of the photovoltaic power generation device is predicted based on the actual value of the output of the photovoltaic power generation device. This configuration makes it possible to collect learning data necessary for creating a prediction model, thereby improving the accuracy of predictions when the first planning method is executed.

[0083] (Technology 9) A control device comprising: a memory that stores a predicted value of power demand of a power consumer and a predicted value of output of a photovoltaic power generation device; and a planner that executes a planning method for planning the output of a fuel cell device so as to compensate for the difference between the predicted value of power demand of the power consumer and the predicted value of output of the photovoltaic power generation device.

[0084] (Technology 10) A power system including a solar power generation device, a fuel cell device, and the control device according to Technology 9.

[0085] The technology of the present disclosure is useful for power systems including photovoltaic power generation devices and fuel cell devices, and is particularly useful for power systems that are distributed power sources.

[0086] REFERENCE SIGNS LIST 10 Control device 12 Prediction unit 14 Planning unit 16 Control unit 20 Database 21 Terminal 22 Weather information service 30 Photovoltaic power generation device 31 Solar cell module 32 First PCS 33 First wattmeter 34 First control device 40 Fuel cell device 41 Fuel cell unit 42 Second PCS 43 Second wattmeter 44 Second control device 50 Power storage device 51 Storage battery 52 Third PCS 53 Third wattmeter 54 Third control device 100 Commercial power supply 102 Power line 200 Power system 300 Load 302 Fourth wattmeter

Claims

1. A control method comprising a step of executing a first planning method for planning output of a fuel cell device so as to compensate for a difference between a predicted value of power demand of a power consumer and a predicted value of output of a solar power generation device.

2. A control method as described in claim 1, which corrects the planned value of the output of the fuel cell device based on the storage rate of the storage device predicted from the difference between the actual value of the power demand of the power consumer and the sum of the actual value of the output of the solar power generation device and the planned value of the output of the fuel cell device.

3. A control method as claimed in claim 1 or 2, in which an upper limit value for the output of the solar power generation device is set to a value less than the difference between the sum of the actual value of the demand power of the power consumer and the chargeable power of the storage device which is less than the maximum charging power of the storage device, and the actual value of the output of the fuel cell device.

4. The control method according to claim 1, further comprising controlling the power storage device so as to compensate for a difference between an actual value of the power demand of the power consumer and the sum of an actual value of the output of the solar power generation device and an actual value of the output of the fuel cell device by charging or discharging the power storage device.

5. The control method of claim 1, wherein the predicted value of the power demand of the power consumer is a value predicted based on a correlation between actual weather conditions in the area where the power consumer is located and the actual value of the power demand of the power consumer, and a weather forecast for the area where the power consumer is located, and is corrected based on the actual value of the power demand of the power consumer in a most recent first period and the predicted value of the power demand of the power consumer in a second period corresponding to the first period.

6. The control method of claim 5, wherein the power consumer includes a factory, and the predicted value of the power demand of the power consumer is a value predicted based on a correlation between the operation record of the factory, the weather record, and the actual value of the power demand of the power consumer, the weather forecast, and the operation plan of the factory, and corrected based on the actual value of the power demand of the power consumer in the first period and the predicted value of the power demand of the power consumer in the second period.

7. A control method as described in claim 1, wherein the predicted value of the output of the solar power generation device is a value predicted based on a correlation between actual weather conditions in the area where the power consumer is located and the actual value of the output of the solar power generation device, and a weather forecast for the area where the power consumer is located, and is corrected based on the actual value of the output of the solar power generation device for the most recent third period and the predicted value of the output of the solar power generation device for a fourth period corresponding to the third period.

8. A control method as described in claim 1, further comprising the steps of: executing, before the first planning method, a second planning method for planning the output of the fuel cell device so as to compensate for the difference between an actual value of the power demand of the power consumer and an actual value of the output of the solar power generation device; and executing the second planning method for a predetermined period of time and then switching to the first planning method, wherein a predicted value of the power demand of the power consumer is predicted based on the actual value of the power demand of the power consumer, and a predicted value of the output of the solar power generation device is predicted based on the actual value of the output of the solar power generation device.

9. A control device comprising: a memory device that stores a predicted value of power demand of an electric power consumer and a predicted value of output of a solar power generation device; and a planner that executes a planning method for planning the output of a fuel cell device so as to compensate for the difference between the predicted value of power demand of the electric power consumer and the predicted value of output of the solar power generation device.

10. A power system comprising: a solar power generation device; a fuel cell device; and the control device according to claim 9.

Images