Charging control method and charging control device

Abstract

This charging control method includes predicting a load power amount to be used by a load (40) during a management period, calculating a prediction accuracy for the predicted load power amount, controlling charging of a charging element (25) during the management period by using a first mode if the prediction accuracy is equal to or higher than a mode determination threshold value, and controlling the charging of the charging element (25) in a second mode in which the charging power amount of the charging element (25) during the management period is smaller than that in the first mode if the prediction accuracy is lower than the mode determination threshold value.

Description

Charging Control Method and Charging Control Device

[0001] The present invention relates to a charging control method and a charging control device.

[0002] Patent Document 1 discloses a control device for a power storage device in a consumer that is connected to a power system and includes a plurality of load devices and at least one power storage device. In the control device, a long-term prediction circuit predicts the temporal change in the power consumption of the consumer as long-term predicted power using a long-term prediction model. A short-term prediction circuit predicts the temporal change in the power consumption of the consumer over a time period immediately after the current time as short-term predicted power based on the temporal change in the power consumption of the consumer over a time period immediately before the current time using a short-term prediction model. The control circuit controls the charging and discharging of the power storage device so as to set a predetermined charging power or a predetermined discharging power based on the long-term predicted power, and controls the discharging of the power storage device so as to set a predetermined discharging power based on the short-term predicted power.

[0003] Japanese Patent No. 6920658

[0004] However, in the method disclosed in Patent Document 1, when there is a deviation between the predicted power and the power actually used, the amount of power used during the time period may exceed the target amount of power.

[0005] An object of the present invention is to provide a charging control method and a charging control device that can control power demand within the range of a target amount of power by appropriately controlling the charging of a charging element.

[0006] A charging control method according to an aspect of the present invention is executed by a charging control device that controls the charging of a charging element with a management period as a control unit in a power system that supplies power to a power demand facility including a load and a charging element. This charging control method includes predicting the amount of load power used by the load during the management period, calculating the prediction accuracy for the amount of load power, and controlling the charging of the charging element in a second mode in which the amount of charging power of the charging element during the management period is small when the prediction accuracy is smaller than a mode determination threshold.

[0007] According to one aspect of the present invention, by appropriately controlling the charging of a charging element, power demand can be controlled within the range of a target power amount. This makes it possible to keep the amount of power consumed by a power demand facility during the management period below the target power amount.

[0008] Figure 1 is a schematic diagram showing the configuration of the power system according to this embodiment. Figure 2 is a diagram illustrating the demand forecasting model. Figure 3 is a diagram illustrating the relationship between facility power consumption and target power consumption. Figure 4 is a flowchart relating to charging control according to this embodiment. Figure 5 is a diagram showing a specific example of charging control according to this embodiment. Figure 6 is a diagram illustrating charging control in the use-up mode. Figure 7 is a diagram illustrating charging control in the save mode. Figure 8 is a diagram showing a modified example of charging control according to this embodiment.

[0009] The following describes a power system to which the charging control device according to this embodiment is applied, and a charging control method, with reference to the drawings.

[0010] As shown in Figure 1, the power system 1 is a system for managing power supply and demand facilities such as commercial facilities including residential facilities, shops and offices (either one or both), public facilities, and factories. The power system 1 includes a power management device 10, a charger 20, a load 40, and renewable energy power generation equipment 50. The charger 20, load 40, and renewable energy power generation equipment 50 are installed in the power supply and demand facilities. Power is supplied to the power supply and demand facilities from the power grid 2.

[0011] Power System 2 is a system that controls the flow of electricity from both the supply and demand sides, and can optimize the flow of electricity. Power System 2 is a concept that includes smart grids, smart communities, and microgrids or MEMS (Mansion Energy Management Systems) that manage energy from the supply source to the consumption part via a communication network within a limited area such as a business or factory.

[0012] The power grid 2 is connected to the charger 20 via the power meter 3, and is also connected to the load 40 and the renewable energy power generation equipment 50 via the power meter 3. The power meter 3 measures the electricity used in the power supply and demand facilities.

[0013] The power management device 10 controls the power supplied from the power system 2 to the load 40 and the electric vehicle 25 connected via the charger 20. The power management device 10 includes a controller composed of a general-purpose electronic circuit including a CPU and peripheral devices such as memory. A computer program is installed in the power management device 10, and the controller executes the computer program, thereby enabling the power management device 10 to perform various power control functions.

[0014] The charger 20 is equipment for charging the electric vehicle 25. The charger 20 is connected to the electric vehicle 25, for example, by a power connector. The charger 20 can also discharge power from the electric vehicle 25 connected to it. The charger 20 may be installed outside the electric vehicle 25, or it may be inside the electric vehicle 25. In this embodiment, one electric vehicle 25 is exemplified, but there may be multiple electric vehicles 25. The electric vehicle 25 is an example of a "charging element" that charges power. Charging elements broadly include elements equipped with batteries, such as automobiles, motorcycles, railway vehicles, play equipment, tools, home appliances, and daily necessities. Charging elements also include stationary batteries.

[0015] Load 40 includes one or more devices that use electricity in a power supply facility. Load 40 is supplied with power as load power, which is the power required for load 40 to operate.

[0016] The renewable energy power generation facility 50 is equipment for generating electricity using renewable energy and supplying the generated electricity. The renewable energy power generation facility 50 is, for example, a solar power generation facility that generates electricity using sunlight. However, power generation using renewable energy is not limited to solar power generation; it may also be wind power generation, geothermal power generation, or other power generation methods that utilize renewable energy. In the power system 1, the electricity generated by the renewable energy power generation facility 50 can be supplied to the electric vehicle 25 and load 40 in place of, or together with, the electricity from the power grid 2.

[0017] The details of the power management device 10 are described below. The power management device 10 controls the power supplied from the power system 2 to the load 40, and also controls the power supplied from the power system 2 to the electric vehicle 25. The power management device 10 communicates with the power meter 3 to obtain the load power used by the load 40. The power management device 10 can communicate with the facility management device 30 to obtain information such as the number of visitors to the power supply facility and weather conditions, and can also communicate with the renewable energy power generation equipment 50 to obtain information on the amount of power generated by the renewable energy power generation equipment 50.

[0018] One of the features of this embodiment is that the power management device 10 also functions as a charging control device that controls the charging of the electric vehicle 25 at predetermined management periods. In this case, the power management device 10 controls the charging of the electric vehicle 25 so that the sum of the load energy used by the load 40 from the power system 2 and the amount of charging energy that the electric vehicle 25 receives from the power system 2 during the management period (hereinafter referred to as "facility energy") is less than or equal to a predetermined target energy amount. Here, the management period is a period used as a control unit by the power management device 10, and corresponds to the period used for calculating charges for the power system 2. For example, in calculating charges for the power system 2, the amount of electricity used is measured every hour from 0 to 30 minutes and every hour from 30 to 60 minutes, and the usage charge is calculated based on this amount of electricity used. The power management device 10 controls the charging of the electric vehicle 25 in accordance with the period for measuring the amount of electricity used in the power system 2. Note that the management period does not necessarily have to be 30 minutes, as long as it follows the measurement period for the amount of electricity used in the power system 2. Furthermore, the target power quantity is the amount of electricity that can be used by the power-demanding facility during the management period. Although the target power quantity is a target value for the amount of electricity in accordance with the contract with power grid 2, the power-demanding facility is not prevented from using more electricity than the target power quantity.

[0019] In this embodiment, the power management device 10 predicts the power demand during the management period at the start timing of the management period (0 minutes and 30 minutes of each hour). That is, the power management device 10 predicts the amount of load power used by the load 40 from the start to the end of the management period. As shown in Figure 2, the power management device 10 has a demand forecasting model 15 for predicting the amount of load power.

[0020] The demand forecasting model 15 takes inputs such as the date and time, the number of visitors to the power demand facility, the amount of power generated by the renewable energy power generation facility 50, the predicted load power amount during the previous management period, the actual load power amount used during the previous management period, and weather conditions including weather and temperature, and outputs a predicted value for load power. The demand forecasting model 15 is composed of a learning model generated by machine learning using historical load power data, historical number of visitors to the power demand facility, historical power generation data from the renewable energy power generation facility 50, and historical weather conditions as training data.

[0021] Incidentally, the load power used by load 40 changes due to various factors, so the actual load power may not be as predicted. Therefore, in addition to the load power, the demand forecasting model 15 outputs the prediction accuracy of the load power predicted by the demand forecasting model 15. Factors that affect the prediction accuracy include the time of day, the number of visitors to the power demand facility, and the amount of power generated by the renewable energy power generation facility 50.

[0022] For example, if the power demand facility is an office building, the time employees arrive at the office may vary from day to day. For instance, on one day employees might arrive around 8:30 a.m., but on another day they might not arrive until after 9:00 a.m. In this case, there will be variations in the time when the load power starts up from day to day. Therefore, the prediction accuracy will be lower during times when the daily variation in load power does not fall within a certain range.

[0023] Furthermore, an increase in the number of visitors to the power demand facility increases the utilization rate of load 40, which in turn increases the load power. In other words, the load power is highly dependent on the number of visitors to the power demand facility. The prediction of load power takes into account the number of visitors to the power demand facility that is expected based on past performance, but if the actual number of visitors deviates significantly from the expected number, the prediction accuracy will be low. However, if the actual number of visitors is less than the expected number, the likelihood of the load power exceeding the target power is low, so in this embodiment, it may be considered that the prediction accuracy is low only when the actual number of visitors is greater than the expected number.

[0024] Furthermore, the amount of electricity generated by the renewable energy power generation facility 50 is affected by weather conditions. For example, in the case of a solar power generation facility, if sunlight is blocked by clouds, the amount of electricity generated will decrease. When the amount of electricity generated by the renewable energy power generation facility 50 is large, if the amount of electricity generated temporarily decreases due to weather conditions, the amount of electricity supplied from the power grid 2 to meet the electricity demand of the load 40 will increase significantly. For this reason, when the amount of electricity generated by the renewable energy power generation facility 50 is large, the load electricity amount may fluctuate greatly, resulting in lower prediction accuracy.

[0025] Therefore, in generating the demand forecasting model 15, learning is performed so that the above factors are reflected. Specifically, past load power performance is compared, and time periods when the daily variation of load power does not fall within a certain range are identified. Then, the demand forecasting model 15 is trained to calculate a lower forecast accuracy during the identified time periods. At this time, the demand forecasting model 15 may be trained to output a lower forecast accuracy the greater the daily variation of load power.

[0026] Similarly, the demand forecasting model 15 is trained to calculate lower prediction accuracy as the actual number of visitors to the power demand facility increases compared to the assumed number of visitors to the power demand facility, for example, the average number of visitors over time in the past. Furthermore, the demand forecasting model 15 is trained to calculate lower prediction accuracy as the amount of electricity generated by the renewable energy power generation facility 50 increases.

[0027] In addition, the prediction accuracy may also be calculated by referring to the results of the previous management period. For example, if the actual load power deviated significantly from the predicted load power in the previous management period, there is a high probability that the actual load power will also deviate significantly from the predicted load power in the current management period that follows the previous one. Therefore, the demand forecasting model 15 may be trained to output a lower prediction accuracy the greater the discrepancy between the predicted load power and the actual load power in the previous management period.

[0028] In the example shown in Figure 2, the demand forecasting model 15 is configured to output load power and forecast accuracy, respectively. However, the demand forecasting model 15 may also output load power as a value with a certain range. In this case, the power management device 10 may treat the magnitude of the range of load power as the forecast accuracy. That is, the power management device 10 may calculate that the larger the range of load power, the lower the forecast accuracy, and that the smaller the range of load power, the higher the forecast accuracy.

[0029] Next, with reference to Figure 3, the concept of charge control according to this embodiment will be explained. The upper part of Figure 3 shows the changes in predicted load power Pe (W) and charge power Pc (W) during each management period from 8:30 to 9:00 and from 9:00 to 9:30. In Figure 3, the charge power Pc is drawn on top of the predicted load power Pe. The target power is the power that the power demand facility can use during the management period, and corresponds to the value obtained by dividing the target power by the management period. The lower part of Figure 3 shows the changes in load energy Ae (Wh), which is the integrated value of the predicted load power Pe, and charge energy Ac (Wh), which is the integrated value of the charge power Pc, during each management period from 8:30 to 9:00 and from 9:00 to 9:30. In Figure 3, the charge energy Ac is drawn on top of the predicted load energy Ae.

[0030] If the predicted load power Ae during the management period is less than the target power, the difference can be allocated to the charging power Ac. In this case, at 9:00, the facility power, obtained by adding the charging power Ac to the predicted load power Ae, will match the target power. Therefore, if the load power actually used by load 40 is the same as the predicted load power Ae, the facility power can be kept below the target power while ensuring the maximum charging power Ac.

[0031] On the other hand, during the management period from 9:00 to 9:30, suppose that after a certain time, the actual load power Pa exceeds the predicted load power Pe. In this case, at 9:30, the facility power amount, which is the sum of the actual load power amount Aa and the charging power amount Ac, exceeds the target power amount. In other words, if the electric vehicle 25 is charged throughout the entire management period, the facility power amount will exceed the target power amount. As a countermeasure, it is conceivable to set the charging power Pc to zero when the actual load power Pa becomes greater than the predicted load power Pe. However, if the discrepancy between the predicted load power Pe and the actual load power Pa is large, the facility power amount may still exceed the target power amount at 9:30. In other words, if the electric vehicle 25 is charged from the start of the management period, if the actual load power deviates significantly from the predicted load power, it will become impossible to keep the facility power amount below the target power amount.

[0032] Therefore, in this embodiment, the power management device 10 has two control modes for controlling the charging of the electric vehicle 25. The first control mode is a use-out mode in which the electric vehicle 25 is charged from the start of the management period so as to use up the amount of power obtained by subtracting the load power amount from the target power amount, i.e., the surplus power amount. However, when the prediction accuracy is low, when controlled in use-out mode, the facility power amount may exceed the target power amount due to an increase in load power demand in the latter half of the management period. Therefore, the power management device 10 has a second control mode, a save mode in which the amount of power charged to the electric vehicle 25 during the management period is smaller than that in the use-out mode. This save mode is a mode that suppresses charging to the electric vehicle 25 in the initial stage of the management period so as to be able to cope even if the power demand increases midway through the management period. Based on these two control modes, the power management device 10 controls the charging of the electric vehicle 25 during the management period using the use-out mode when the prediction accuracy is equal to or greater than the mode determination threshold. On the other hand, if the prediction accuracy is lower than the mode determination threshold, the power management device 10 controls the charging of the electric vehicle 25 using the save mode.

[0033] The charging control method according to this embodiment will now be described with reference to Figure 4. The flowchart shown in Figure 4 is executed by the power management device 10, triggered at 0 and 30 minutes past every hour.

[0034] The power management device 10 predicts the load power, which is the amount of electricity used by the load 40 during the management period (S10). The power management device 10 also calculates the prediction accuracy of the load power (S11). As described above, the power management device 10 uses the demand forecasting model 15 to predict the load power and calculate the prediction accuracy.

[0035] If the prediction accuracy is equal to or greater than the mode determination threshold (S12: YES), the power management device 10 controls the charging of the electric vehicle 25 in the use-out mode (S13). On the other hand, if the prediction accuracy is less than the mode determination threshold (S12: NO), the power management device 10 controls the charging of the electric vehicle 25 in the save mode (S14).

[0036] The charging control for each control mode will be explained with reference to Figures 5 to 7. Figure 5 shows the prediction accuracy, control mode, load power, and charging energy amount for each management period from 8:00 to 10:00.

[0037] In the management period starting at 8:00 and the management period starting at 9:30, the prediction accuracy is above the mode determination threshold, so the power management device 10 selects the use-up mode. In the charge control of the use-up mode, as shown in Figure 6, the power management device 10 calculates the differential power by subtracting the load power actually used by the load from the target power at a predetermined control cycle. Then, the power management device 10 determines the amount of power to charge the electric vehicle 25 by integrating the calculated differential power. The power management device 10 controls the charging of the electric vehicle 25 in accordance with the determined amount of power to charge. As shown in Figure 5, in the use-up mode, the surplus power obtained by subtracting the load power from the target power is allocated as the amount of power to charge the electric vehicle 25, and the electric vehicle 25 is charged in conjunction with the start of the management period. In the management period starting at 8:00, since the load power remains constant during the management period, the amount of power to charge also increases linearly. In the management period starting at 9:30, the load power decreased towards the latter half of the management period, resulting in a larger increase in the amount of power being charged towards the latter half of the management period.

[0038] In contrast, for the management period starting at 8:30 and the management period starting at 9:00, the prediction accuracy is lower than the mode determination threshold, so the power management device 10 selects save mode. In save mode, the power management device 10 sets the amount of charge to zero until a first timing t1, which is a certain amount of time (for example, 15 minutes) after the start of the management period. Then, from the first timing t1 to the end of the management period, the power management device 10 allocates the amount of power that would be surplus even if the load 40 were operating at maximum power as the amount of charge for the electric vehicle 25. Specifically, as shown in Figure 7(a), the power management device 10 refers to a pre-held map or calculation formula to identify the maximum load power A when the load 40 is operating at maximum power between the first timing t1 and the end of the management period. Also, as shown in Figure 7(b), the power management device 10 identifies the cumulative power B, which is the cumulative load power used from the start of the management period to the first timing t1. Then, as shown in Figure 7(c), the power management device 10 determines the minimum surplus power amount D, which is obtained by subtracting the sum of the maximum load power amount A and the cumulative power B from the facility target power amount, as the power amount to charge the electric vehicle 25. The power management device 10 controls the charging of the electric vehicle 25 in accordance with the determined power amount. As shown in Figure 5, in save mode, the power amount to charge is zero from the start of the management period until the first timing t1. Then, considering the load power amount (cumulative power B) from the start of the management period until the first timing t1 and the load power amount (maximum load power amount A) when the load 40 operates at maximum power from the first timing t1 until the end of the management period, the minimum surplus power amount (minimum surplus power amount) is allocated to charge the electric vehicle 25. In the management period starting at 8:30, the load power up to the first timing t1 is small, so even if the load 40 operates at maximum power in the latter half of the management period, there is surplus power. Therefore, charging of the electric vehicle 25 occurs in the latter half of the management period. On the other hand, in the management period starting at 9:00, the load power is high up to the first timing t1, so even if the load power decreases after the first timing t1, charging of the electric vehicle 25 does not occur in the latter half of the management period.

[0039] As described above, the charging control method according to this embodiment includes predicting the amount of load power used by the load 40 during the management period, calculating the prediction accuracy for the predicted amount of load power, controlling the charging of the charging element during the management period using the first mode if the prediction accuracy is equal to or greater than a mode determination threshold, and controlling the charging of the charging element in the second mode, in which the amount of charge of the charging element during the management period is smaller than that of the first mode, if the prediction accuracy is less than the mode determination threshold.

[0040] According to this method, when the accuracy of load power prediction is low, charging control is performed so that the amount of power charged by the charging element during the management period is smaller compared to when the accuracy of load power prediction is high. Even in cases where the actual load power exceeds the predicted load power, it is possible to prevent the load power during the management period from exceeding the target amount of power. In this way, by appropriately controlling the charging of the electric vehicle 25, the power demand can be controlled within the range of the target amount of power, and the facility's power consumption can be kept below the target amount of power.

[0041] In the charging control method according to this embodiment, the first mode controls the charging of the electric vehicle 25 during the management period based on the difference power obtained by subtracting the load power actually used by the load 40 from the target power usable by the power demand facility. In contrast, the second mode controls the charging of the electric vehicle 25 during the management period based on the minimum surplus power obtained by subtracting the sum of the load power amount from the start of the management period to the first timing t1 and the maximum load power amount from the first timing t1 to the end of the management period from the target power amount during the management period.

[0042] According to this method, in the exhaustion mode, the electric vehicle 25 can be charged from the start of the management period so as to exhaust the amount of electric power obtained by subtracting the load electric power from the target amount of electric power, that is, the surplus amount of electric power. When the prediction accuracy of the load power is high, there is a high possibility that the predicted load power amount and the actual load power amount will match. In this case, by selecting the exhaustion mode, it is possible to secure the maximum amount of charging power while keeping the facility power amount below the target power amount. Also, in the save mode, charging of the electric vehicle 25 is suppressed at the initial stage of the management period. If there is a margin in the amount of electric power, considering the maximum load power, the minimum surplus amount of electric power (minimum surplus power) will be allocated to the charging power amount of the electric vehicle 25. When the prediction accuracy of the load power is low, a case where the actual load power exceeds the predicted load power is likely to occur. In this case, by selecting the save mode, it is possible to charge the electric vehicle 25 while appropriately using the surplus amount of electric power while keeping the facility power amount below the target power amount.

[0043] In the charging control method according to the present embodiment, when the management period corresponds to a time zone in which the daily variation of the load power used by the load does not fall within a certain range, the prediction accuracy is calculated to be low. In the time zone where the load power varies daily, it becomes difficult to predict the load power amount. According to the method of the present embodiment, since the tendency of the use of the load power in the power demand facility can be taken into account, the prediction accuracy can be appropriately calculated.

[0044] In the charging control method according to the present embodiment, the more the number of visitors to the power demand facility actually visiting the facility exceeds the number of visitors to the power demand facility assumed in the management period, the lower the prediction accuracy is calculated. In the prediction of the load power amount, the number of visitors to the power demand facility assumed from past performance is taken into account. However, when the actual number of visitors deviates from the assumed number of visitors, it becomes difficult to predict the load power amount. According to the method of the present embodiment, since the number of visitors affecting the transition of the load power can be taken into account, the prediction accuracy can be appropriately calculated.

[0045] In the charging control method according to this embodiment, the power demand facility further includes a renewable energy power generation facility 50 that generates power using renewable energy. In this case, the greater the power generation amount by the renewable energy power generation facility 50, the lower the prediction accuracy is calculated. When the power generation amount of the renewable energy power generation facility 50 is large, when the power generation amount temporarily decreases due to weather conditions, the power supplied from the power grid 2 to meet the power demand of the load 40 increases significantly. According to the method of this embodiment, since the power generation amount by the renewable energy power generation facility that affects the transition of the load power can be taken into account, the prediction accuracy can be appropriately calculated.

[0046] In the charging control method according to this embodiment, the greater the deviation between the predicted load power amount in the previous management period and the actual load power amount in the previous management period, the lower the prediction accuracy is calculated in the current management period. When the actual load power amount deviates significantly from the predicted load power amount in the previous management period, there is a high possibility that the actual load power amount will also deviate significantly from the predicted load power amount in the current management period following the previous one. According to the method of this embodiment, since the performance of the previous management period can be taken into account, the prediction accuracy can be appropriately calculated.

[0047] Note that in this embodiment, the prediction accuracy is calculated at the start of the management period. However, as shown in FIG. 8, the power management device 10 may recalculate the prediction accuracy at the second timing t2 in the middle of the management period. And when the current control mode is the save mode, if the prediction accuracy calculated at the second timing t2 is higher than the mode determination threshold value, the power management device 10 may switch from the save mode to the use-up mode (for example, the management period from 8:30 to 9:00).

[0048] According to this method, when it can be determined that the prediction accuracy has improved in the middle of the management period, the control mode is changed to the use-up mode. Thereby, while increasing the charging power amount, the facility power amount can be kept below the target power amount.

[0049] Furthermore, if the power management device 10 is controlling the charging power of the electric vehicle 25 in use-out mode and the load power changes by more than a predetermined value, it may switch from use-out mode to save mode (for example, during a management period from 9:30 to 10:00).

[0050] Even with high prediction accuracy, large changes in actual load power can cause the prediction to deviate from the actual value. Therefore, switching to save mode allows the facility's power consumption to be kept below the target level.

[0051] One or more embodiments of the present invention are not limited to the charging control method described above, but also include a charging control device. This charging control device is a power system that supplies power to a power demand facility comprising a power-consuming load 40 and an electric vehicle 25 capable of charging power, and includes a controller that controls the charging of the electric vehicle with a predetermined management period as the control unit. This controller executes the charging control method described above. With this device, when the accuracy of load power prediction is low, charging control is performed so that the amount of power charged by the charging element during the management period is smaller compared to when the accuracy of load power prediction is high. Even in cases where the actual load power exceeds the predicted load power, it is possible to suppress the load power during the management period from exceeding the target amount of power. In this way, by appropriately controlling the charging of the electric vehicle 25, power demand can be controlled within the range of the target amount of power, and the facility's power consumption can be kept below the target amount of power.

[0052] As described above, embodiments of the present invention have been presented, but the statements and drawings that constitute part of this disclosure should not be understood as limiting the invention. Various alternative embodiments, examples, and operational techniques will become apparent to those skilled in the art from this disclosure.

[0053] 1: Power system, 2: Power grid, 20: Charger, 25: Electric vehicle, 30: Facility management device, 40: Load, 50: Renewable energy power generation equipment

Claims

1. A power system that supplies power to a power demand facility comprising a power-consuming load and a charging element capable of charging the power, wherein a charging control device controls the charging of the charging element with a predetermined management period as the control unit, comprising: predicting the amount of load power used by the load during the management period; calculating the prediction accuracy for the amount of load power; controlling the charging of the charging element during the management period using a first mode if the prediction accuracy is equal to or greater than a mode determination threshold; and controlling the charging of the charging element during the management period using a second mode in which the amount of power charged by the charging element during the management period is less than that of the first mode if the prediction accuracy is less than the mode determination threshold.

2. The charging control method according to claim 1, wherein the first mode controls the charging of the charging element during the management period based on the differential power obtained by subtracting the load power actually used by the load from the target power usable by the power demand facility, and the second mode controls the charging of the charging element during the management period based on the minimum surplus power obtained by subtracting the sum of the integrated power obtained by accumulating the load power from the start of the management period to a first timing and the maximum load power when the load operates at maximum power from the first timing to the end of the management period from the target power usable by the power demand facility during the management period.

3. The charging control method according to claim 1 or 2, wherein the prediction accuracy is calculated to be lower if the management period falls within a time period in which the daily variation of the load power used by the load does not fall within a certain range.

4. The charging control method according to claim 1 or 2, wherein the prediction accuracy is calculated to be lower if the number of visitors actually visiting the power demand facility is greater than the number of visitors expected to visit the power demand facility during the management period.

5. The charging control method according to claim 1 or 2, wherein the power demand facility further comprises a renewable energy power generation facility that generates electricity using renewable energy, and the prediction accuracy is calculated to be lower the larger the amount of electricity generated by the renewable energy power generation facility.

6. The charging control method according to claim 1 or 2, wherein the greater the discrepancy between the predicted load power amount in the previous management period and the actual load power amount used by the load in the previous management period, the lower the prediction accuracy for the current management period is calculated.

7. The charging control method according to any one of claims 1 to 6, wherein, when the charging of the charging element is controlled in the second mode, the prediction accuracy of the load power amount is recalculated in the middle of the management period, and if the prediction accuracy calculated in the middle of the management period becomes higher than the mode determination threshold, the method switches from the second mode to the first mode.

8. A charging control method according to any one of claims 1 to 7, wherein when the charging of the charging element is controlled in the first mode, if the load power actually used by the load changes by more than a predetermined value, the method switches from the first mode to the second mode.

9. A power system that supplies power to a power demand facility comprising a load that uses power and a charging element capable of charging said power, comprising a controller that controls the charging of the charging element with a predetermined management period as the control unit, wherein the controller predicts the amount of load power used by the load during the management period, calculates the prediction accuracy for the amount of load power, controls the charging of the charging element during the management period using a first mode if the prediction accuracy is equal to or greater than a mode determination threshold, and controls the charging of the charging element during the management period using a second mode in which the amount of power charged by the charging element during the management period is less than that of the first mode if the prediction accuracy is less than the mode determination threshold.

Images