Load control device, load control program, and load control method

Abstract

To provide a load control device, a load control program, and a load control method that accurately perform demand adjustment with a simple configuration.SOLUTION: A regression equation creation unit 105 creates a regression equation that represents a demand of an air-conditioning apparatus 21 as a temperature-sensitive demand on the basis of past actual data on total demand. An air-conditioning apparatus demand estimation unit 102 obtains total demand data and estimates a current demand of the air-conditioning apparatus 21 using the regression equation created by the regression equation creation unit 105. A demand control unit 104 calculates a predicted value for the total demand at a predetermined time and controls the air-conditioning apparatus 21 so that the total demand at the predetermined time falls within a predetermined demand target value on the basis of the predicted value and the current demand of the air-conditioning apparatus 21 estimated by the air-conditioning apparatus demand estimation unit 102.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a load control device, a load control program, and a load control method.

Background Art

[0002] In recent years, due to changes in the global environment and social issue environment, not only large-scale power plants but also the utilization of renewable energy such as solar power generation has been actively promoted. On the other hand, renewable energy is greatly affected by nature such as weather and temperature, and there is also a problem that the influence of fluctuations on the power generation amount cannot be avoided. This fluctuation amount has been increasing due to the recent expansion of the introduction of renewable energy. Therefore, during peak demand periods, the power supply and demand becomes tight, while during low demand periods, the supply becomes excessive, and a situation where electricity from renewable energy remains surplus also occurs.

[0003] In order to avoid such a situation, in order to change the demand according to the power generation amount of renewable energy and stabilize the demand balance, demand adjustment using consumer equipment has been studied. When considering countermeasures using consumer equipment, it is expected that a business will be established in which intermediaries and retail electricity suppliers called aggregators that utilize consumer-side resources contribute to the demand adjustment of the grid operator and obtain compensation.

[0004] As one method of demand adjustment using consumer equipment, it is conceivable to utilize demand response (DR). Demand response is a technology for adjusting demand by providing economic benefits. For example, as an application of demand response, demand adjustment such as controlling the battery of a consumer by giving incentives or penalties for the charge and discharge of the battery in the consumer is conceivable.

[0005] In this type of demand adjustment, retail electricity providers and aggregators support the adjustment of each customer's demand so as not to exceed the demand target value, which is the expected maximum value of the amount of demand from each customer, in order to meet the customer's requests for energy conservation and excess reduction. For example, one way to support customers in energy conservation and excess reduction is to place sensors that measure the output of each load device used by the customer, such as air conditioners and lighting equipment, and then control each load device using the measurement information from each sensor.

[0006] In addition, as a demand adjustment technique, a technology has been proposed that receives energy usage values ​​and outdoor temperature values ​​from a building, uses regression analysis to determine the air conditioning coefficient and non-air conditioning coefficient based on the temperature difference between the reference temperature and the outdoor temperature value and the amount of energy used, and then controls air conditioning equipment. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] U.S. Patent Application Publication No. 2011 / 0106471 [Overview of the project] [Problems that the invention aims to solve]

[0008] However, using sensors to measure the output of load equipment in order to appropriately control the demand at individual customers is complicated and requires a lot of work, such as installing a sensor for each load piece of equipment, making it difficult to achieve accurate demand adjustment. Furthermore, in technologies that control load equipment by calculating air conditioning coefficients etc. using regression analysis based on temperature differences and energy consumption, the estimated value of each load piece of equipment is allocated based on the demand ratio to estimate the demand for each, but simple allocation may deviate from actual demand, making it difficult to achieve accurate demand adjustment.

[0009] The disclosed technology was developed in view of the above, and aims to provide a load control device, a load control program, and a load control method that accurately adjust demand with a simple configuration. [Means for solving the problem]

[0010] In one embodiment of the load control device, load control program, and load control method disclosed in this application, a regression equation creation unit creates a regression equation that represents the demand for air conditioning equipment as temperature-sensitive demand based on historical total demand data. An air conditioning equipment demand estimation unit acquires total demand data and estimates the current demand for air conditioning equipment using the regression equation created by the regression equation creation unit. A demand control unit calculates a predicted value of total demand at a predetermined time and controls the air conditioning equipment based on the predicted value and the current demand for air conditioning equipment estimated by the air conditioning equipment demand estimation unit so that the total demand at the predetermined time falls within a predetermined demand target value. [Effects of the Invention]

[0011] In one respect, the present invention enables accurate demand adjustment with a simple configuration. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 is a schematic diagram of the load control system. [Figure 2] Figure 2 is a block diagram of the load control device. [Figure 3] Figure 3 shows an example of the relationship between temperature and total demand for a specific customer. [Figure 4] Figure 4 shows an example of cooling variables. [Figure 5] Figure 5 shows an example of a regression equation when classified according to temperature and operating conditions. [Figure 6] Figure 6 is a diagram illustrating the method for allocating aggregate demand data. [Figure 7] Figure 7 shows an example of the results of the demand estimation for air conditioning equipment. [Figure 8]FIG. 8 is a diagram showing a comparison of demand estimation results between the load control device according to Example 1 and the method of apportioning by demand ratio. [Figure 9] FIG. 9 is a diagram for explaining a method of calculating the amount of demand reduction of air conditioning equipment. [Figure 10] FIG. 10 is a flowchart of demand adjustment processing by the load control device according to Example 1. [Figure 11] FIG. 11 is a flowchart of demand estimation processing by the load control device according to Example 1. [Figure 12] FIG. 12 is a flowchart of load control processing by the load control device according to Example 1. [Figure 13] FIG. 13 is a diagram for explaining the classification according to the operating status according to Example 2. [Figure 14] FIG. 14 is a diagram for explaining a more detailed classification according to the operating status on high-operation days and low-operation days. [Figure 15] FIG. 15 is a frequency distribution diagram of regression errors. [Figure 16] FIG. 16 is a diagram showing the duration curve of regression errors. [Figure 17] FIG. 17 is a hardware configuration diagram of the load control device.

Embodiments for Carrying Out the Invention

[0013] Hereinafter, embodiments of the load control device, load control program, and load control method disclosed in the present application will be described in detail based on the drawings. Note that the load control device, load control program, and load control method disclosed in the present application are not limited by the following embodiments.

Examples

[0014] FIG. 1 is a schematic diagram of a load control system. The load control system is arranged as part of the power system of Customer 1. Customer 1 is an enterprise or household that uses electricity supplied from Electric Utility 2.

[0015] As shown in Figure 1, the load control system includes a load control device 10, a meter for transactions 11, and a pulse meter 12. The load control system also includes various load devices that consume power, such as air conditioning equipment 21 and other load devices 22, including lighting equipment and electric water heaters. In Figure 1, only one other load device 22 is shown, but there are usually multiple other load devices 22, and there is no particular limit to their number. The air conditioning equipment 21 and the other load devices 22 use electricity supplied by the electric utility company 2.

[0016] Here, load control refers to a control system that controls the output of customer 1's air conditioning equipment 21 and other load equipment 22, thereby keeping the cumulative demand for customer 1's air conditioning equipment 21 and other load equipment 22 within a demand target value (demand target value).

[0017] The transaction meter 11 is, for example, a smart meter. The transaction meter 11 is connected to the electricity provider 2. The transaction meter 11 receives power supply from the electricity provider 2. The transaction meter 11 then supplies power to other load equipment 22 such as air conditioning equipment 21 and lighting equipment. The transaction meter 11 also measures the amount of electricity used by customer 1 at predetermined intervals, such as the cumulative amount of electricity and the 30-minute average power, and detects changes in customer 1's electricity consumption. The transaction meter 11 is equipped with a pulse meter 12.

[0018] The pulse meter 12 extracts the total and instantaneous output values ​​of all load equipment, including the air conditioning equipment 21 and other load equipment 22, as pulse data. The pulse meter 12 then outputs the extracted pulse data to the load control device 10.

[0019] The load control device 10 may be installed in, for example, an Energy Management System (EMS). An EMS is a system that visualizes demand, which is the energy usage status, and adjusts the operation of various load devices to optimize demand. The load control device 10 is connected to the pulse meter 12. The load control device 10 is also connected to the air conditioning equipment 21 and the other load devices 22, respectively.

[0020] The load control device 10 controls the cumulative demand of all load equipment of customer 1 by reducing the output of controlled equipment such as air conditioning equipment 21 by outputting control command values ​​to the controlled equipment. For example, if the demand target value is given as the cumulative demand value over 30 minutes, the load control device 10 calculates the control amount every 30 minutes and outputs a control command value. However, the timing of outputting the control command value is not limited to this and may be at other times.

[0021] Figure 2 is a block diagram of the load control device. The load control device 10 according to Embodiment 1, as shown in Figure 2, includes an excess determination unit 101, an air conditioning equipment demand estimation unit 102, an other load equipment demand estimation unit 103, a demand control unit 104, a regression equation creation unit 105, and a database 106. Hereinafter, the process of controlling when the target value of demand is exceeded will be referred to as the demand adjustment process.

[0022] Database 106 stores historical total demand data for customer 1, organized daily and corresponding to the temperature for that day. Database 106 may be located on a different device from the load control device 10.

[0023] The regression equation generation unit 105 pre-generates a regression equation used to calculate the total demand for other load equipment 22, which is the sum of the demand for air conditioning equipment 21 and the demand for each other load equipment 22, based on historical daily total demand data stored in the database 106. The method of generating the regression equation by the regression equation generation unit 105 is described below.

[0024] This section explains the basic principle of the demand for air conditioning equipment 21, which is expressed as temperature-sensitive demand. Figure 3 shows an example of the relationship between temperature and total demand at a specific customer. In Figure 3, the horizontal axis represents temperature, and the vertical axis represents the cumulative value of total demand over 30 minutes. Figure 3 is an example where a demand target value is obtained every 30 minutes. That is, 48 ​​graphs like the one shown in Figure 3 can be created in a day. Line segment 200 in Figure 3 shows the change in total demand according to temperature. The area below line segment 200 on the paper represents the total demand at customer 1, which is the sum of the demand for air conditioning equipment 21 and the demand for other load equipment 22.

[0025] As shown in Figure 3, total demand is classified into three types: cooling demand 201, heating demand 202, and other demand 203. Cooling demand 201 is the demand when the air conditioning equipment 21 is cooling, and this demand increases as the temperature rises. Heating demand 202 is the demand when the air conditioning equipment 21 is heating, and this demand increases as the temperature falls. Other demand 203 is the demand in total demand other than cooling demand 201 and heating demand 202, and is the total demand for other load equipment 22. Other demand 203 is constant regardless of the temperature and represents the demand when the air conditioning equipment 21 is not operating.

[0026] For example, the relationship between temperature and total demand, as shown in Figure 3, can be considered to differ for each individual customer when classifying changes in demand in detail. Furthermore, when classifying changes in demand broadly, the relationship between temperature and total demand can be considered to differ depending on the type of facility, such as residential or commercial, and even further, depending on the industry within a commercial facility. In other words, the graph shown in Figure 3 can be generated for each classification according to the classification of individual customers. In the following explanation, assuming that changes in demand differ for each individual customer, the graph in Figure 3 will be created for each individual customer. Also, the relationship between temperature and total demand differs from time to time. In other words, the graph shown in Figure 3 can be created for each hour.

[0027] When estimating the demand for air conditioning equipment 21, the demand for air conditioning equipment 21 can be extracted as temperature-sensitive demand by finding a regression equation representing line segment 200 in Figure 3. Therefore, the regression equation creation unit 105 creates the regression equation for line segment 200 as follows. Here, we will explain using cooling demand as an example.

[0028] In this embodiment 1, a cooling variable Xc, obtained by converting actual temperature data, is used as a variable to represent the use of cooling demand at a specific temperature. The cooling variable Xc can be expressed by the following formula (1). (1) Xc = max(0, T - Cth)

[0029] Here, max is a function that takes the larger of 0 and the temperature minus the cooling threshold value. T represents the temperature. Cth is the temperature at which the use of the air conditioning equipment 21 for cooling demand begins, and is called the cooling threshold. Figure 4 is a diagram showing an example of the cooling variables. In Figure 4, the horizontal axis represents the temperature, and the vertical axis represents the cumulative value of the total demand over a 30-minute period from a given time. Figure 4 is an example of a graph that enlarges the portion of the cooling demand 201 in Figure 3. Line segment 210 represents the cooling variables and corresponds to the range that includes the region of cooling demand 201 of line segment 200. In range 211, T-Cth is less than 0, so the cooling variable Xc = 0. Also, in range 212, T-Cth ≥ 0, so the cooling variable Xc = T-Cth.

[0030] The regression equation generation unit 105 selects the value that best fits the regression equation to the actual demand and temperature data as the cooling threshold Cth. For example, the regression equation generation unit 105 searches in 1°C increments, such as 0°C, 1°C, 2°C, ..., to find the cooling threshold Cth.

[0031] The regression equation generation unit 105 applies the regression equation represented by the following formula (2) to the actual demand and temperature data to determine the regression equation for each time period. (2) yt = βc × Xc + β

[0032] Here, yt represents the demand for each time period t, which is obtained by dividing the day into 30-minute intervals. βc is the regression coefficient used to calculate demand, and β is a constant term used to adjust the value. Since the regression equation is created separately for each time period, the constant term and regression coefficient change with time period t.

[0033] Furthermore, the following two conditions can be assumed regarding fluctuations in demand. One is that the demand for air conditioning equipment 21 changes depending on its operating status and temperature. The other is that the demand for other load equipment 22 changes depending on its operating status.

[0034] Therefore, under these two assumed conditions, in order to extract temperature-sensitive demand with higher accuracy, the regression equation creation unit 105 classifies each day based on actual demand and temperature data and creates a regression equation for each classification. Figure 5 shows an example of a regression equation when classified according to temperature and operating status. In Figure 5, the horizontal axis represents temperature, and the vertical axis represents the cumulative value of total demand for 30 minutes from a predetermined time. Figure 5 shows an example in which each day is classified into six groups. It is preferable to select a number of groups that corresponds to the classification of heating and cooling demand at the target customer 1. For example, if the heating and cooling demand at customer 1 differs between holidays and weekdays, each day can be divided into two groups. Also, if the heating and cooling demand differs depending on the day of the week, each day can be divided into seven groups, and if the operating status of Saturdays and Sundays is similar, it can be divided into six groups.

[0035] The regression equation generation unit 105 has a predetermined number of groups to classify and a ranking of the groups according to the operating status. For example, the regression equation generation unit 105 has a ranking of the groups, with the group with the lowest demand value being the 1st group, and the groups being the 2nd, 3rd, 4th, 5th, and 6th groups as the demand increases.

[0036] The regression equation generation unit 105 plots points representing the actual demand and temperature data of customer 1 on the coordinate space, as shown by the points in Figure 5. Next, the regression equation generation unit 105 classifies each plotted point into a predetermined number of groups in 1°C increments, for example, 0°C, 1°C, 2°C, ... In this case, the regression equation generation unit 105 can perform the classification using clustering methods or the like. However, the regression equation generation unit 105 may use other classification methods as long as they can classify according to the relationship between demand and temperature.

[0037] Next, the regression equation creation unit 105 connects the temperature-classified groups according to their ranking to generate a series of consecutive groups based on temperature. Then, using formulas (1) and (2), the regression equation creation unit 105 selects a cooling threshold Cth for each group and adapts the regression equation to create a regression equation that represents the total demand for each group. In this embodiment 1, the regression equation creation unit 105 creates a linearly represented regression equation using, for example, the least squares method. However, the regression equation does not have to be linear, and a quadratic curve or the like may be used. In this way, the regression equation creation unit 105 creates regression equations for each customer 1, categorized by the operating status of the air conditioning equipment 21 and other load equipment 22, and by time of day. The regression equation creation unit 105 creates and stores these regression equations in advance.

[0038] As described above, the regression equation creation unit 105 creates a regression equation that represents the demand for air conditioning equipment 21 as temperature-sensitive demand based on historical total demand data. More specifically, the regression equation creation unit 105 classifies and groups the historical total demand data according to the operating status of air conditioning equipment 21 and other load equipment 22, and generates a regression equation for each group. In this process, the regression equation creation unit 105 classifies the historical total demand data by clustering.

[0039] The excess determination unit 101 has a demand target value, which is a reference value that the total demand at customer 1 should be kept within that value. The excess determination unit 101 can, for example, use the demand target value as the demand target value. For example, if the demand target value is given as the cumulative demand value over 30 minutes, the excess determination unit 101 will use the demand target value given every 30 minutes as the demand target value.

[0040] The excess detection unit 101 receives pulse data input from the pulse meter 12 according to the amount of usage. The excess detection unit 101 then obtains total demand data for all load equipment, including the air conditioning equipment 21 and other load equipment 22, from the acquired pulse data. For example, if the demand target value is the cumulative demand value every 30 minutes, the excess detection unit 101 obtains the cumulative value of the total demand over 30 minutes as total demand data.

[0041] The excess determination unit 101 then determines whether the total demand, as shown in the acquired total demand data, exceeds the demand target value. If the total demand does not exceed the demand target value, the excess determination unit 101 terminates the demand adjustment process. On the other hand, if the total demand exceeds the demand target value, the excess determination unit 101 outputs the total demand data to the air conditioning equipment demand estimation unit 102 and the demand control unit 104.

[0042] The air conditioning equipment demand estimation unit 102 receives total demand data from the excess determination unit 101. The air conditioning equipment demand estimation unit 102 also acquires temperature data from the thermometer 13. Furthermore, the air conditioning equipment demand estimation unit 102 acquires a regression equation from the regression equation creation unit 105 that corresponds to the current operating status of the air conditioning equipment 21 and other load equipment 22 and the current time period.

[0043] The air conditioning equipment demand estimation unit 102 then uses the regression equation, temperature, and total demand data to calculate the current demand for air conditioning equipment 21 as a temperature-sensitive demand that changes according to the temperature. Here, the air conditioning equipment demand estimation unit 102 estimates the current demand for air conditioning equipment 21, but hereafter it will simply be referred to as "demand for air conditioning equipment 21". The following describes the process by which the air conditioning equipment demand estimation unit 102 estimates the demand for air conditioning equipment 21.

[0044] The air conditioning equipment demand estimation unit 102 substitutes the current temperature obtained from the thermometer 13 into the acquired regression equation to calculate the standard demand value for air conditioning equipment 21, such as cooling demand or heating demand, corresponding to the current temperature. The air conditioning equipment demand estimation unit 102 also calculates the total standard demand value for other load equipment 22 corresponding to the current temperature. However, since the total standard demand value for other load equipment 22 is constant, the air conditioning equipment demand estimation unit 102 will obtain a constant value.

[0045] Next, the air conditioning equipment demand estimation unit 102 uses the calculated standard demand value for air conditioning equipment 21 and the total standard demand value for other load equipment 22 to apportion the measured total demand data between air conditioning equipment 21 and other load equipment 22, thereby calculating the demand for air conditioning equipment 21 and the total demand for other load equipment 22.

[0046] Figure 6 is a diagram illustrating the method of allocating total demand data. The horizontal axis of Figure 6 represents temperature, and the vertical axis represents the cumulative total demand over a 30-minute period starting from a predetermined time. In Figure 6, demand 231 corresponds to the standard demand value for air conditioning equipment 21 when the temperature is T°C. Also, demand 232 corresponds to the standard total demand value for other load equipment 22 when the temperature is T°C. Here, the standard demand value for air conditioning equipment 21 is denoted as ysa, and the standard total demand value for other load equipment 22 is denoted as yso.

[0047] For example, the air conditioning equipment demand estimation unit 102 calculates the regression error ε as the difference between the total demand data and the sum of the calculated demand standard value for air conditioning equipment 21 and the total demand standard value for other load equipment 22. That is, in Figure 6, the total demand data for T℃ is represented by point 233, so the air conditioning equipment demand estimation unit 102 calculates the demand 234 obtained by subtracting demands 231 and 232 from the value of point 233 as the regression error ε. Next, the air conditioning equipment demand estimation unit 102 uses the following formulas (3) and (4) to apportion the regression error ε by the ratio of the demand standard value ysa for air conditioning equipment 21 and the total demand standard value yso for other load equipment 22 to calculate the demand Pa for air conditioning equipment 21 and the total demand Po for other load equipment 22. (3) Pa=ysa+ε×(ysa / (ysa+yso)) (4) Po=yso+ε×(yso / (ysa+yso))

[0048] The air conditioning equipment demand estimation unit 102 can similarly calculate the demand for air conditioning equipment 21 and the total demand for other load equipment 22 in both cases of cooling and heating demand for air conditioning equipment 21. The air conditioning equipment demand estimation unit 102 then outputs the calculated information on the demand for air conditioning equipment 21 and the total demand for other load equipment 22 to the other load equipment demand estimation unit 103. The air conditioning equipment demand estimation unit 102 also outputs the calculated information on the demand for air conditioning equipment 21 to the demand control unit 104.

[0049] Figure 7 shows an example of the results of air conditioning equipment demand estimation. Using the method described above, graphs 310, 320, and 330 in Figure 7 represent the estimated results of the demand for air conditioning equipment 21 calculated by the air conditioning equipment demand estimation unit 102. Graphs 310, 320, and 330 represent 30-minute time periods on the horizontal axis and the cumulative total demand for 30 minutes from a predetermined time on the vertical axis. Region 311 of graph 310 represents the estimated results of heating demand, region 321 of graph 320 and region 331 of graph 330 represent the estimated results of cooling demand, and the remaining regions represent other demands. By connecting the temperature-sensitive demand for one year, including heating and cooling demand, as shown in graphs 310, 320, and 330 extracted by time period, time-series data for temperature-sensitive demand over 365 days × 48 time periods is obtained.

[0050] As described above, the air conditioning equipment demand estimation unit 102 acquires total demand data and estimates the current demand for air conditioning equipment 21 using the regression equation created by the regression equation creation unit 105. The air conditioning equipment demand estimation unit 102 also calculates the standard demand value for air conditioning equipment 21 and the total standard demand value for other load equipment 22 based on the regression equation, calculates the regression error of the regression equation, and estimates the current demand for air conditioning equipment 21 by apportioning the calculated regression error. In this case, the air conditioning equipment demand estimation unit 102 apportions the regression error by the ratio of the standard demand value for air conditioning equipment 21 to the total standard demand value for other load equipment 22.

[0051] Returning to Figure 2, the explanation continues. The other load equipment demand estimation unit 103 has a demand ratio for each other load equipment 22 relative to the total demand data estimated from past performance data of the demand for each other load equipment 22. For example, if there are three other load equipment 22, the other load equipment demand estimation unit 103 has a demand ratio of 25:15:10.

[0052] The Other Load Equipment Demand Estimation Unit 103 receives information on the total demand for other load equipment 22 from the Air Conditioning Equipment Demand Estimation Unit 102. The Other Load Equipment Demand Estimation Unit 103 then allocates the total demand for other load equipment 22 according to the demand ratio to calculate the demand for each of the other load equipment 22. After that, the Other Load Equipment Demand Estimation Unit 103 outputs the information on the demand for each of the other load equipment 22 to the Demand Control Unit 104.

[0053] Figure 8 shows a comparison of the demand estimation results between the load control device according to Example 1 and the method of apportionment by demand ratio. Table 341 shows the demand estimation results using the load control device 10 according to Example 1. Table 342 shows the demand estimation results when demand estimation is performed by apportionment by demand ratio.

[0054] The load control device 10 according to this embodiment 1 extracts and estimates the demand for air conditioning equipment 21 as temperature-sensitive demand. That is, since the demand for air conditioning equipment 21 is considered to increase as the temperature rises or falls, the load control device 10 according to this embodiment 1 improves the accuracy of the demand estimation for air conditioning equipment 21 compared to simply allocating by demand ratio. For example, as shown in Table 341, the load control device 10 according to this embodiment 1 matches the true value of the demand for air conditioning equipment 21, but in the method of allocating by demand ratio, there is a large error between the estimated value and the true value of the demand for air conditioning equipment 21. Furthermore, since the demand for air conditioning equipment 21 accounts for a large proportion of the total demand, improving the accuracy of the estimate of the demand for air conditioning equipment 21 also improves the accuracy of the estimate of total demand. For example, in Table 342, the cumulative miscalculation between the estimated value and the true value of total demand is 80 kWh. In contrast, in Table 341, the cumulative miscalculation between the estimated value and the true value of total demand is 20 kWh. Thus, it can be seen that the load control device 10 according to this embodiment 1 shows improved accuracy in estimating demand compared to the case where demand estimation was performed by apportionment based on the demand ratio.

[0055] Returning to Figure 2, the explanation continues. The demand control unit 104 receives demand information for the air conditioning equipment 21 from the air conditioning equipment demand estimation unit 102. The demand control unit 104 also receives demand information for each of the other load devices 22 from the other load device demand estimation unit 103. The demand control unit 104 also has a target value for demand.

[0056] Next, the demand control unit 104 calculates a forecast value for the total demand during the time period to be adjusted. For example, the demand control unit 104 may use the total demand for the same time period on the same day in the past as the forecast value. Alternatively, the demand control unit 104 may use the cumulative demand from 0 to 15 minutes and the demand at the 15-minute mark to forecast the cumulative demand from 15 to 30 minutes. The demand control unit 104 may use any method for forecasting total demand.

[0057] Next, the demand control unit 104 calculates the target excess amount Pe by subtracting the target demand value from the predicted total demand value. Then, the demand control unit 104 checks whether the target excess amount Pe is greater than 0 and determines whether the predicted total demand value exceeds the target demand value.

[0058] If the forecast value of total demand exceeds the target value of demand, the demand control unit 104 calculates the amount of reduction in demand for the air conditioning equipment 21, Pamax, based on the temperature setting adjustment limit. Specifically, the demand control unit 104 calculates the amount of reduction in demand for the air conditioning equipment 21, Pmax, as follows.

[0059] The temperature setting adjustment limit is determined based on a temperature setting that does not disrupt operations, and represents how much the temperature can be changed from the current setting, i.e., the temperature that can be changed from a setting that does not consider demand reduction. For cooling demand, the temperature setting adjustment limit represents how many degrees higher the temperature can be set from the current setting, and for heating demand, it represents how many degrees lower the temperature can be set from the current setting. For example, in the case of cooling demand, if the current set temperature is 23℃ and the temperature can be raised to 25℃, the temperature setting adjustment limit is 2℃. This temperature setting adjustment limit should differ for each customer and preferably be determined individually for each customer. For example, if customer 1 is highly energy-conscious, the temperature setting adjustment limit can be set higher.

[0060] Figure 9 is a diagram illustrating the method for calculating the amount of reduction possible in the demand for air conditioning equipment. In Figure 9, the horizontal axis represents the cooling variable, and the vertical axis represents the total demand for a 30-minute period starting from a given time. Here, we will explain using cooling demand as an example. Let the temperature setting adjustment limit be x°C. The demand for air conditioning equipment 21 with a temperature setting x°C higher corresponds to the demand for air conditioning equipment 21 when the temperature is x°C lower and demand reduction is not a consideration. Also, demand 241 represents the total demand before control, and demand 242 represents the total demand after control.

[0061] Here, if the cooling variable before control is Xc1, then the cooling variable after control is Xc2 = Xc1 - x. Then, if we represent the demand for the air conditioning equipment 21 before control as Pa1, the total demand for other load equipment 22 before control as Po1, the demand for the air conditioning equipment 21 after control as Pa2, and the total demand for other load equipment 22 after control as Po2, then the demand control unit 104 can calculate Pa1, Pa2, Po1, and Po2 using the following equations (5) to (8). Here, equations (5) and (6) correspond to equations (3) and (4). (5) Pa1=ysa+ε×(ysa / (ysa+yso)) (6) Po1=yso+ε×(yso / (ysa+yso)) (7) Pa2 = Pa1 × ((Xc1 - x) / Xc1) (8) Po2 = Po1

[0062] Therefore, the demand control unit 104 calculates the amount of reduction in the demand for the air conditioning equipment 21, Pamax, using the following formula (9). Although cooling demand has been used as an example here, the demand control unit 104 can calculate the amount of reduction in heating demand, Pamax, in the same way. (9) Pamax = Pa1 - Pa2 =ysa+ε×(ysa / (ysa+yso))-Pa1×((Xc1-x) / Xc1) =(ysa+ε×ysa / (ysa+yso))×(1-(Xc1-x) / Xc1) =(ysa+ε×ysa / (ysa+yso))×(x / Xc1)

[0063] Next, the demand control unit 104 determines whether the calculated amount of reduction in the demand for the air conditioning equipment 21 is greater than the target excess. If the calculated amount of reduction in the demand for the air conditioning equipment 21 is greater than the target excess, the demand control unit 104 determines the target excess as the amount of reduction in the demand for the air conditioning equipment 21. In this case, the demand control unit 104 does not need to reduce the demand for other load equipment 22. Then, the demand control unit 104 controls the air conditioning equipment 21 so that the determined amount of reduction in the demand for the air conditioning equipment 21 is reduced.

[0064] In response to this, if the calculated amount of reduction in the demand for the air conditioning equipment 21 is less than or equal to the target excess amount, the demand control unit 104 determines the amount of reduction in the demand for the air conditioning equipment 21 as the amount of reduction in the demand for the air conditioning equipment 21. Next, the demand control unit 104 selects other load equipment 22 that will not interfere with operations. For example, the demand control unit 104 can select other load equipment 22 specified by customer 1 as other load equipment 22 that will not interfere with operations.

[0065] Next, the demand control unit 104 calculates the total reduction in demand for other load equipment 22 by subtracting the amount of reduction in demand for air conditioning equipment 21 from the target excess amount. Here, the demand control unit 104 has a demand ratio for each other load equipment 22 relative to the total demand data estimated from the historical demand data of each other load equipment 22. The demand control unit 104 calculates the demand reduction amount for each of the selected other load equipment 22 by apportioning the total reduction in demand for the other load equipment 22 according to the demand ratio of the selected other load equipment 22.

[0066] The demand control unit 104 then controls the air conditioning equipment 21 so that the demand reduction amount determined from the demand of the air conditioning equipment 21 is reduced. The demand control unit 104 also controls the selected other load equipment 22 so that the demand reduction amount determined from the demand of each of the selected other load equipment 22 is reduced.

[0067] As described above, the demand control unit 104 calculates a predicted value of the total demand at a predetermined time, and controls the air conditioning equipment 21 based on the predicted value and the current demand of the air conditioning equipment 21 estimated by the air conditioning equipment demand estimation unit 102, so that the total demand at the predetermined time falls within a predetermined demand target value. More specifically, the demand control unit 104 calculates the amount that can reduce the demand for the air conditioning equipment 21, and based on the amount that can be reduced, determines whether the total demand at the predetermined time will fall within the demand target value by controlling the air conditioning equipment 21. If the total demand at the predetermined time will fall within the demand target value by controlling the air conditioning equipment 21, the unit controls the air conditioning equipment 21. If the total demand at the predetermined time will not fall within the demand target value by controlling the air conditioning equipment 21, the unit controls the air conditioning equipment 21 and other load equipment 22 to bring the total demand at the predetermined time within the demand target value. Furthermore, the demand control unit 104 has a pre-defined temperature setting adjustment limit for the air conditioning equipment 21, which is the temperature difference that allows the set temperature to be changed. Based on the current demand and temperature difference of the air conditioning equipment 21 estimated by the air conditioning equipment demand estimation unit 102, the demand control unit 104 calculates the amount by which the demand for the air conditioning equipment 21 can be reduced. However, as is clear from the modified equation in formula (9), the demand control unit 104 can calculate the amount by which the demand for the air conditioning equipment 21 can be reduced without directly using the predicted value of the total demand for the air conditioning equipment 21.

[0068] Figure 10 is a flowchart of the demand adjustment process by the load control device according to Example 1. Next, the flow of the demand adjustment process by the load control device 10 according to Example 1 will be explained with reference to Figure 10.

[0069] The excess determination unit 101 receives pulse data input from the pulse meter 12 according to the amount of usage. The excess determination unit 101 then obtains total demand data from the pulse data, which shows the total demand for all load equipment, including the air conditioning equipment 21 and other load equipment 22 (step S1).

[0070] Next, the excess determination unit 101 determines whether the total demand shown in the total demand data has exceeded a predetermined demand target value (step S2). If the total demand has not exceeded the demand target value (step S2: negative), the excess determination unit 101 terminates the demand adjustment process.

[0071] In response to this, if the total demand exceeds the demand target value (Step S2: Affirmative), the excess determination unit 101 outputs the total demand data to the air conditioning equipment demand estimation unit 102 and the demand control unit 104. The air conditioning equipment demand estimation unit 102 estimates the demand for air conditioning equipment 21 using the current temperature, a pre-created regression equation, and the total demand data (Step S3). At this time, the air conditioning equipment demand estimation unit 102 also estimates the total demand for other load equipment 22.

[0072] The other load equipment demand estimation unit 103 estimates the demand for each of the other load equipment 22 by apportioning the total demand for other load equipment 22 estimated by the air conditioning equipment demand estimation unit 102 according to the proportion of demand held (step S4).

[0073] The demand control unit 104 calculates a predicted value for the total demand during the time period to be adjusted. Then, using the predicted total demand, the target demand value, the temperature setting adjustment limit, the estimated demand for the air conditioning equipment 21, and the estimated demand for other load equipment 22, it determines the equipment to be controlled, including the air conditioning equipment 21, and further determines the amount of demand reduction for each piece of equipment to be controlled (step S5).

[0074] Next, the demand control unit 104 controls each of the controlled devices according to the determined demand reduction amount (step S6).

[0075] Subsequently, the demand control unit 104 determines whether or not to terminate the control based on whether or not a control termination instruction has been received from customer 1 (step S7). If the control is to be continued (step S7: negative), the demand adjustment process returns to step S1. Conversely, if the control is to be terminated (step S7: positive), the demand control unit 104 terminates the demand adjustment process.

[0076] Figure 11 is a flowchart of the demand estimation process by the load control device according to Embodiment 1. Next, the flow of the demand estimation process by the load control device 10 according to Embodiment 1 will be explained with reference to Figure 11. Each process in the flow of Figure 11 is an example of the process executed in step S3 of the flow of Figure 10.

[0077] The air conditioning equipment demand estimation unit 102 uses regression equations created in advance by the regression equation creation unit 105, which represent heating demand and cooling demand as temperature-sensitive demand, to calculate the demand reference value for air conditioning equipment 21 and the total demand reference value for other load equipment 22 corresponding to the current temperature (step S101).

[0078] Next, the air conditioning equipment demand estimation unit 102 calculates the regression error by subtracting the demand standard value for air conditioning equipment 21 and the total demand standard value for other load equipment 22 from the measured total demand obtained from the total demand data (step S102).

[0079] Next, the air conditioning equipment demand estimation unit 102 calculates the demand for air conditioning equipment 21 and the total demand for other load equipment 22 using formulas (3) and (4). Specifically, the air conditioning equipment demand estimation unit 102 apportions the regression error by the ratio of the standard demand value for air conditioning equipment 21 to the standard demand value for other load equipment 22. Then, the air conditioning equipment demand estimation unit 102 adds each of the apportionment results to the standard demand value for air conditioning equipment 21 and the standard demand value for other load equipment 22 to calculate the demand for air conditioning equipment 21 and the total demand for other load equipment 22, and uses these as estimated values ​​(step S103).

[0080] Figure 12 is a flowchart of the load control process by the load control device according to Embodiment 1. Next, the flow of the load control process by the load control device 10 according to Embodiment 1 will be explained with reference to Figure 12. Each process in the flow of Figure 12 is an example of the process executed in step S5 of the flow of Figure 10.

[0081] The demand control unit 104 calculates a forecast value of the total demand for the time period to be adjusted. Next, the demand control unit 104 calculates the target excess amount Pe by subtracting the target demand value from the forecast value of total demand (step S201).

[0082] Then, the demand control unit 104 determines whether the target excess amount Pe is greater than 0 (Pe>0) (step S202).

[0083] If the target excess amount Pe is 0 or less (step S202: negation), the forecast value of total demand is within the target value of demand, so the demand control unit 104 terminates the load control process.

[0084] In contrast, if the target excess amount Pe is greater than 0 (step S202: affirmative), that is, if the predicted total demand exceeds the target demand, the demand control unit 104 calculates the amount of reduction Pamax of the demand for the air conditioning equipment 21 (step S203). Specifically, the demand control unit 104 can calculate the amount of reduction Pamax using formula (9), which is obtained from the demand reference value of the air conditioning equipment 21, the total demand reference value of other load equipment 22, the regression error, the adjustment limit of the temperature setting, and the cooling coefficient or heating coefficient before control obtained from the current temperature.

[0085] Next, the demand control unit 104 determines whether the calculated amount of reduction in the demand for the air conditioning equipment 21, Pamax, is greater than the target excess amount Pe (Pamax > Pe) (step S204).

[0086] If the amount of reduction in demand for the air conditioning equipment 21 Pamax is greater than the target excess amount Pe (step S204: affirmative), the demand control unit 104 determines the target excess amount Pe as the amount of reduction in demand for the air conditioning equipment 21 Par (Par = Pe) (step S205).

[0087] The demand control unit 104 controls the air conditioning equipment 21 so that the demand reduction amount determined for the demand of the air conditioning equipment 21 is reduced (step S206).

[0088] In contrast, if the amount of reduction in demand for the air conditioning equipment 21 Pamax is less than or equal to the target excess amount Pe (step S204: negation), the demand control unit 104 determines the amount of reduction in demand for the air conditioning equipment 21 Pamax to be the amount of reduction in demand for the air conditioning equipment 21 Par (Pamax = Par) (step S207).

[0089] Next, the demand control unit 104 selects an alternative load device 22 that will not interfere with operations (step S208).

[0090] Next, the demand control unit 104 subtracts the amount of reduction possible in the demand for the air conditioning equipment 21 from the target excess amount to calculate the total reduction in demand for other load equipment 22, Por (Por = Pe - Pamax) (step S209).

[0091] Next, the demand control unit 104 calculates the amount of demand reduction for each of the selected other load devices 22 by apportioning the total demand reduction amount Por of the other load devices 22 according to the demand ratio of the other load devices 22 (step S210).

[0092] The demand control unit 104 then controls the air conditioning equipment 21 so that the demand reduction amount determined from the demand of the air conditioning equipment 21 is reduced. The demand control unit 104 also controls the selected other load equipment 22 so that the demand reduction amount determined from the demand of each of the selected other load equipment 22 is reduced (step S211).

[0093] As described above, the load control device according to this embodiment 1 estimates the demand for air conditioning equipment and the demand for load equipment other than air conditioning equipment according to temperature using a regression equation created in advance based on total demand data. Furthermore, the load control device predicts the demand at the time to be controlled, and if the predicted demand exceeds the demand target value, it calculates the amount by which the demand for air conditioning equipment can be reduced from the total demand data. If the demand target value can be achieved by reducing the demand for air conditioning equipment, the load control device controls the air conditioning equipment to achieve the demand target value. If the reduction in the demand for air conditioning equipment is insufficient, the load control device controls the air conditioning equipment and other load equipment to reduce the demand for air conditioning equipment to the amount by which it can be reduced, as well as to reduce the demand for other load equipment that does not interfere with operations, in order to achieve the demand target.

[0094] This allows for the control of various load devices, including the customer's air conditioning equipment, to achieve demand targets based on pulse data from trading meters, without the need to install sensors for output measurement on each load device. Therefore, accurate demand adjustment can be performed with a simple configuration. [Examples]

[0095] Next, Example 2 will be described. The load control device 10 according to Example 2 is also represented by the block diagram in Figure 2. The load control device 10 according to Example 2 differs from that of Example 1 in its method of classifying the operating status of the air conditioning equipment 21 and other load equipment 22. In the following description, the operation of each part, which is the same as in Example 1, will be omitted.

[0096] In Example 1, in order to extract temperature-sensitive demand with high accuracy, points representing the temperature and demand for each day were plotted on a graph and classified according to the operating status of each day by clustering. In contrast, in Example 2, the regression equation creation unit 105 performs the classification according to the operating status of each day by the following method.

[0097] Figure 13 is a diagram illustrating the classification by operating status according to Example 2. Figure 14 is a diagram illustrating a more detailed classification by operating status on high-operating days and low-operating days.

[0098] The regression equation generation unit 105 calculates the total daily demand (kWh / day) for each day. Next, the regression equation generation unit 105 creates the frequency distribution diagram 401 shown in Figure 13. The frequency distribution diagram 401 represents the total daily demand on the horizontal axis and the frequency on the vertical axis. The frequency distribution diagram 401 is a diagram that summarizes the data for a one-month period.

[0099] The regression equation generation unit 105 then divides the frequency distribution diagram 401 into regions according to the characteristics of the operating conditions and performs classification according to the operating conditions. For example, in the case of the frequency distribution diagram 401, the region can be divided at the minimum point, and the regression equation generation unit 105 divides the frequency distribution diagram 401 into three regions: regions 411 to 413. In this case, region 411 corresponds to high operating days, region 412 corresponds to low operating days, and region 413 corresponds to non-operating days. Here, the number of classifications may be other than three patterns depending on the characteristics of the frequency distribution.

[0100] Graph 402 shows the operating patterns for each of the operating conditions represented by regions 411 to 413. In Graph 402, the horizontal axis represents time, and the vertical axis represents cumulative demand over 30 minutes. Graph 402 shows the average value of the cumulative demand over 30 minutes for each day in each of regions 411 to 413. Curve 421 in Graph 402 is the operating pattern on a high-operating day in region 411. Curve 422 is the operating pattern on a low-operating day in region 412. Curve 423 is the operating pattern on a non-operating day in region 413. As can be seen from Graph 402, there are distinctive operating patterns for the air conditioning equipment 21 and other load equipment 22 in each of regions 411 to 413. It is clear that demand is high on high-operating days in region 411, lower on low-operating days in region 412 than on high-operating days, and almost no demand on non-operating days in region 413. Therefore, it can be confirmed that the classification of operating status, divided into regions 411 to 413 by the air conditioning equipment demand estimation unit 102, is correct.

[0101] Furthermore, in Example 2, the regression equation creation unit 105 performs a more detailed classification of high-operating days and low-operating days based on the characteristics of their operating status. Specifically, the regression equation creation unit 105 creates the frequency distribution diagram 403 shown in Figure 14 for high-operating days and low-operating days. In the frequency distribution diagram 403, the horizontal axis represents the operating hours per day, and the vertical axis represents the frequency. The frequency distribution diagram 403 is a diagram that summarizes the data for a one-month period.

[0102] The regression equation generation unit 105 then divides the frequency distribution diagram 403 into regions according to the characteristics of the frequency of demand, as demand changes depending on the operating conditions, and performs classification according to the operating conditions. For example, in the case of the frequency distribution diagram 403, the region can be divided at the local minimum, and the regression equation generation unit 105 divides the frequency distribution diagram 403 into three regions: regions 431 to 433. In this case, region 431 corresponds to long-hour operating days, region 432 corresponds to fixed-hour operating days, and region 433 corresponds to short-hour operating days. Here again, the number of classifications may be other than three patterns depending on the characteristics of the frequency distribution.

[0103] Graph 404 shows the operating patterns for each operating condition represented by regions 431 to 433. In Graph 404, the horizontal axis represents time, and the vertical axis represents cumulative demand over 30 minutes. Curve 441 in Graph 404 represents the operating pattern on long-hour operating days in region 431. Curve 442 represents the operating pattern on regular-hour operating days in region 432. Curve 443 represents the operating pattern on short-hour operating days in region 433. As can be seen from Graph 404, there are distinctive operating patterns for the air conditioning equipment 21 and other load equipment 22 in each of regions 431 to 433. On long-hour operating days in region 431, demand occurs over a long period of time, on regular-hour operating days in region 432, which are normal operating hours, the duration of demand is shorter than on long-hour operating days, and on short-hour operating days in region 433, demand is shorter than on regular-hour operating days. Therefore, it can be confirmed that the classification of operating status, divided into regions 431 to 433 by the air conditioning equipment demand estimation unit 102, is correct.

[0104] Based on the above, the regression equation creation unit 105 can classify each day into seven groups according to its operating status. Subsequently, the regression equation creation unit 105 creates a regression equation for each of the created operating status groups, similar to the example in Example 1. The air conditioning equipment demand estimation unit 102 uses the regression equation created by the regression equation creation unit 105 in the manner described above to estimate the demand for air conditioning equipment 21 and the total demand for other load equipment 22. In this way, the regression equation creation unit 105 classifies historical total demand data based on the frequency of total demand for each day.

[0105] As described above, the load control device according to this embodiment 2 classifies the operating status of air conditioning equipment and load equipment according to the frequency of total demand each day, generates groups according to the operating status, and generates a regression equation representing temperature-sensitive demand for each generated group. This method also makes it possible to extract temperature-sensitive demand with high accuracy. Therefore, even with a method that generates groups according to the operating status of air conditioning equipment and load equipment according to the frequency of total demand each day, it is possible to accurately adjust demand with a simple configuration. [Examples]

[0106] Next, we will describe Example 3. The load control device 10 according to Example 3 is also represented by the block diagram in Figure 2. The load control device 10 according to Example 3 differs from that of Example 1 in its method of allocating regression errors. In the following description, we will omit explanations of the operation of each part, which is the same as in Example 1.

[0107] In Example 1, when allocating the regression error between the demand for the air conditioning equipment 21 and the demand for other load equipment 22, the allocation was performed using the ratio of the demand standard value to the total demand standard value. In contrast, in Example 3, the air conditioning equipment demand estimation unit 102 allocates the regression error using the following method.

[0108] Figure 15 is a frequency distribution diagram of regression errors. In Figure 15, the horizontal axis represents regression errors and the vertical axis represents frequency. Figure 16 is a diagram showing the duration curve of regression errors. In Figure 16, the horizontal axis represents cumulative frequency and the vertical axis represents regression errors.

[0109] The air conditioning equipment demand estimation unit 102 obtains actual data of the total demand for each day during the transitional period when air conditioning equipment 21 is not used, along with temperature information for each day, from the database 106. Next, the air conditioning equipment demand estimation unit 102 calculates the total demand baseline value for each day of other load equipment 22 using a regression equation corresponding to each day. Next, the air conditioning equipment demand estimation unit 102 calculates the regression error, which is the difference between the calculated total demand baseline value and the actual total demand data. Then, the air conditioning equipment demand estimation unit 102 creates a frequency distribution diagram of the regression error for the transitional period, as shown in curve 501 in Figure 15.

[0110] Next, the air conditioning equipment demand estimation unit 102 obtains actual data of the total daily demand for each summer day using the air conditioning equipment 21, along with information on the temperature for each day, from the database 106. Here, we will explain using cooling demand as an example, but the same applies to heating demand. Next, the air conditioning equipment demand estimation unit 102 calculates a total demand standard value by combining the demand standard value for the air conditioning equipment 21 and the total demand standard value for other load equipment 22 for each day using a regression equation corresponding to each day. Next, the air conditioning equipment demand estimation unit 102 calculates the regression error, which is the difference between the calculated total demand standard value and the actual data of total demand. Then, the air conditioning equipment demand estimation unit 102 creates a frequency distribution diagram of the summer regression error for each temperature, as shown in curve 502 in Figure 15. For example, the air conditioning equipment demand estimation unit 102 creates frequency distribution diagrams in 1°C increments, such as ..., 25°C, 26°C, 27°C, ...

[0111] Next, the air conditioning equipment demand estimation unit 102 generates duration curves for the regression errors of curves 501 and 502, respectively, as shown in Figure 16. Curve 511 is the duration curve for the regression error during the intermediate season, represented by curve 501. Curve 512 is the duration curve for the regression error during the summer season, represented by curve 502.

[0112] The air conditioning equipment demand estimation unit 102 then calculates the demand Pa for air conditioning equipment 21 and the total demand Po for other load equipment 22 by apportioning the regression error ε of the current demand, which is the subject of apportionment, using the ratio of a and b in the duration curve of Figure 16. Specifically, the air conditioning equipment demand estimation unit 102 calculates the demand Pa for air conditioning equipment 21 and the total demand Po for other load equipment 22 using the following formulas (10) and (11). The air conditioning equipment demand estimation unit 102 can calculate both cooling demand and heating demand in the same way. (10) Pa = ysa + ε × (a / (a ​​+ b)) (11) Po = yso + ε × (b / (a+b))

[0113] Subsequently, the demand control unit 104 controls the air conditioning equipment 21 and other load equipment 22 so that the total demand falls within the demand target value, using the demand for air conditioning equipment 21 calculated by the air conditioning equipment demand estimation unit 102. In this way, the demand control unit 104 apportions the regression error using the ratio of the regression error between the intermediate period when air conditioning equipment 21 is not operating and the period when air conditioning equipment 21 is operating, based on historical total demand data.

[0114] As explained above, the air conditioning equipment demand estimation unit according to this embodiment 3 allocates the current regression error from the ratio of the regression error between the intermediate period and the period in which air conditioning equipment operates in past total demand data. This method also allows for appropriate allocation of the regression error and enables accurate estimation of the demand for air conditioning equipment and the total demand for other load equipment. Therefore, even with the method of allocating the current regression error from the ratio of the regression error between the intermediate period and the period in which air conditioning equipment operates in past total demand data, it is possible to accurately adjust demand with a simple configuration.

[0115] (modified version) Furthermore, the load control device 10 according to this modified example can also use a calculation method based on the regression error apportionment method in Example 3, which is different from Example 1, as a method for calculating the amount of reduction that can be achieved with the air conditioning equipment 21. The modified example is described below.

[0116] In this case, the demand control unit 104 uses formulas (10) and (11) as formulas for calculating the demand for the air conditioning equipment 21 and the total demand for other load equipment 22.

[0117] For example, let Xc1 be the cooling variable before control, and Xc2 = Xc1 - x be the cooling variable after control. Then, let Pa1' represent the demand for the air conditioning equipment 21 before control, Po1' represent the total demand for other load equipment 22 before control, Pa2' represent the demand for the air conditioning equipment 21 after control, and Po2' represent the total demand for other load equipment 22 after control. Then, the demand control unit 104 can calculate Pa1', Pa2', Po1', and Po2' using the following equations (12) to (15). Here, equations (12) and (13) correspond to equations (10) and (11). (12) Pa1' = ysa + ε × (a / (a ​​+ b)) (13) Po1'=yso+ε×(b / (a+b)) (14) Pa2' = Pa1' × ((Xc1-x) / Xc1) (15) Po2'=Po1'

[0118] Therefore, the demand control unit 104 calculates the amount of reduction in the demand for the air conditioning equipment 21, Pamax, using the following formula (16). Although cooling demand has been used as an example here, the demand control unit 104 can calculate the amount of reduction in heating demand, Pamax, in the same way. (16) Pamax = Pa1' - Pa2' =ysa+ε×(a / (a+b))-Pa1'×((Xc1-x) / Xc1) =(ysa+ε×a / (a+b))×(1-(Xc1-x) / Xc1) =(ysa+ε×a / (a+b))×(x / Xc1)

[0119] By determining the Pamax amount of reduction in the demand for the air conditioning equipment 21 in this way, the demand control unit 104 can calculate the Pamax amount of reduction in the demand for the air conditioning equipment 21 in a manner consistent with the calculation of the demand for the air conditioning equipment 21 by the air conditioning equipment demand estimation unit 102.

[0120] As explained above, the demand control unit in this modified configuration calculates the amount of reduction in the demand for air conditioning equipment that is consistent with the calculation of the demand for air conditioning equipment by the air conditioning equipment demand estimation unit. This method also makes it possible to accurately estimate the amount of reduction in the demand for air conditioning equipment. Therefore, the load control device in this modified configuration can accurately adjust demand with a simple configuration. [Examples]

[0121] Next, Example 4 will be described. The load control device 10 according to Example 4 is also represented by the block diagram in Figure 2. The load control device 10 according to Example 4 differs from Examples 1 and 2 in the method of creating the regression equation.

[0122] The regression equation creation unit 105 in Example 4 creates regression equations using a modified Gaussian Mixture Regression (GMR) method. GMR is a method for fitting multiple regression models to data by assuming that the errors of each regression model follow a normal distribution, when generating multiple regression models with different regression equations as shown in Figure 5. The regression equation creation unit 105 uses the EM algorithm to fit the multiple regression models to the data in a way that maximizes the likelihood, which is an indicator of how well the model fits the data. The EM algorithm consists of calculations called the E step and the M step. The E step corresponds to the step of creating groups according to the operating status, and the M step corresponds to the step of determining the regression equation.

[0123] The regression equation generation unit 105 uses the EM algorithm with the following modifications. In a normal E step, each data point represented in Figure 5 is assigned an estimated probability of which regression model it was generated from. For example, the regression model that generated the first data point has a 30% probability of being the first regression model and a 70% probability of being the second regression model. In contrast, the regression equation generation unit 105 performs the E step by setting the probability of each data point being generated by the regression model with the highest probability to 100% and all others to 0%. The reason for this is that in a normal M step, a weighted least squares method is calculated using this probability as a weight, but by setting this probability to either 0% or 100%, the calculation can be simplified to a normal least squares method.

[0124] Furthermore, the regression equation creation unit 105 executes step M by imposing the following two constraints. The first constraint is that each regression coefficient must be non-negative when determining the regression coefficients of the regression equation using the least squares method. Since the demand for air conditioning equipment 21 and the total demand for other load equipment 22 can both be assumed to be values ​​of 0 or greater, this constraint does not cause any problems. The second constraint is that each regression model is assigned a sequential number in order of increasing demand, and the slope of the regression equation of the i-th regression model in the region where cooling demand and heating demand exist is smaller than that of the (i+1)th regression equation. The effect of this constraint is explained in Figure 5 as preventing the solid lines representing the regression models from intersecting. As a result, the regression equation creation unit 105 can extract temperature-sensitive demand, i.e., create regression equations, under the assumption that no intersections occur between regression equations.

[0125] Here, the EM algorithm is a method that approximately maximizes likelihood, and there is a possibility that it may fall into a local minimum that does not fit the data well, potentially leading to unstable results. In contrast, the regression equation generation unit 105 can suppress the instability of the EM algorithm and avoid falling into a local minimum that does not fit well by imposing the two constraints described above.

[0126] In the manner described above, the regression equation creation unit 105 creates a regression equation using the EM algorithm modified in GMR. The air conditioning equipment demand estimation unit 102 then uses the regression equation created in this manner to estimate the demand for air conditioning equipment 21. In this way, the regression equation creation unit 105 groups past total demand data and creates multiple regression equations by fitting multiple regression equations to past total demand data so that the regression error of each regression equation follows a normal distribution.

[0127] As described above, the load control device according to this embodiment 4 creates a regression equation using a modified EM algorithm in GMR. This makes it possible to extract temperature-sensitive demand with high accuracy and to accurately adjust demand with a simple configuration.

[0128] Here, in each of the above embodiments 1 to 4 and its modifications, a load control device 10 having an excess determination unit 101, an air conditioning equipment demand estimation unit 102, an other load equipment demand estimation unit 103, a demand control unit 104, and a regression equation creation unit 105 has been described. However, even if the current demand does not exceed the demand target value without the excess determination unit 101 performing an excess determination, the air conditioning equipment demand estimation unit 102 may use the regression equation created by the regression equation creation unit 105 to estimate the air conditioning equipment demand, and the demand control unit 104 may use the estimation result to control the air conditioning equipment 21 and other load equipment 22. Furthermore, the demand control unit 104 can calculate the amount of reduction possible in the demand for air conditioning equipment 21 without using the current demand for each other load equipment 22 estimated by the other load equipment demand estimation unit 103, and can control the air conditioning equipment 21 and other load equipment 22. Therefore, the load control device 10 may not have an excess determination unit 101 and an other load equipment demand estimation unit 103. Even with such a configuration, temperature-sensitive demand can be extracted with high accuracy, and demand adjustment can be performed accurately with a simple configuration.

[0129] (Hardware configuration) Figure 17 is a hardware configuration diagram of the load control device. The load control device 10 described in each of the above embodiments 1 to 4 and modified examples has a hardware configuration such as that shown in Figure 17. That is, the load control device 10 has a CPU (Central Processing Unit) 91, memory 92, hard disk 93, and network interface 94. The CPU 91 is connected to the memory 92, hard disk 93, and network interface 94 via a bus.

[0130] The network interface 94 is a communication interface between the load control device 10 and an external device. For example, the network interface 94 relays communication between the pulse meter 12 and the CPU 91.

[0131] The hard disk 93 is an auxiliary storage device. The hard disk 93 can implement the functions of the database 106. The hard disk 93 also stores various programs, including programs for implementing the functions of the excess determination unit 101, the air conditioning equipment demand estimation unit 102, the other load equipment demand estimation unit 103, the demand control unit 104, and the regression equation creation unit 105, as illustrated in Figure 2.

[0132] Memory 92 is the main memory. Memory 92 can use DRAM (Dynamic Random Access Memory).

[0133] The CPU 91 reads various programs from the hard disk 93, loads them into memory 92, and executes them. As a result, the CPU 91 implements the functions of the excess determination unit 101, the air conditioning equipment demand estimation unit 102, the other load equipment demand estimation unit 103, the demand control unit 104, and the regression equation creation unit 105 as illustrated in Figure 2.

[0134] However, as described above, the load control device 10 does not necessarily have an overload determination unit 101 and an other load equipment demand estimation unit 103. In that case, the hard disk 93 stores various programs, including programs for realizing the functions of the air conditioning equipment demand estimation unit 102, the demand control unit 104, and the regression equation creation unit 105, and the CPU 91 realizes the functions of the air conditioning equipment demand estimation unit 102, the demand control unit 104, and the regression equation creation unit 105. [Explanation of Symbols]

[0135] 1 Consumer 2. Electricity Providers 10 Load control device 11. Transaction meter 12. Pulse meter 13 Thermometer 21 Air conditioning equipment 22 Other load equipment 101 Excess judgment section 102 Air Conditioning Equipment Demand Estimation Department 103 Other load equipment demand estimation section 104 Demand Control Unit 105 Regression Equation Creation Section 106 Databases

Claims

1. A regression equation generation unit creates a regression equation that expresses the demand for air conditioning equipment as temperature-sensitive demand based on historical total demand data, An air conditioning equipment demand estimation unit acquires total demand data and estimates the current demand for air conditioning equipment using the regression equation created by the regression equation creation unit, A demand control unit calculates a predicted total demand at a predetermined time and controls the air conditioning equipment based on the predicted value and the current demand for the air conditioning equipment estimated by the air conditioning equipment demand estimation unit, so that the total demand at the predetermined time falls within a predetermined demand target value. A load control device characterized by being equipped with

2. The load control device according to claim 1, characterized in that the demand control unit calculates the amount by which the demand for the air conditioning equipment can be reduced, determines whether the total demand at a predetermined time will fall within the demand target value by controlling the air conditioning equipment based on the amount by which the demand can be reduced, controls the air conditioning equipment if the total demand at a predetermined time will fall within the demand target value by controlling the air conditioning equipment, and controls the air conditioning equipment and other load equipment other than the air conditioning equipment if the total demand at a predetermined time will not fall within the demand target value by controlling the air conditioning equipment.

3. The load control device according to claim 2, characterized in that the demand control unit has a predetermined temperature difference for the air conditioning equipment that allows the set temperature to be changed, and calculates the amount by which the demand for the air conditioning equipment can be reduced based on the current demand for the air conditioning equipment estimated by the air conditioning equipment demand estimation unit and the temperature difference.

4. The load control device according to claim 1, characterized in that the regression equation generation unit classifies and groups past total demand data according to the operating status of the air conditioning equipment and other load equipment other than the air conditioning equipment, and generates the regression equation for each group.

5. The load control device according to claim 4, characterized in that the regression equation generation unit classifies past total demand data by clustering.

6. The load control device according to claim 4, characterized in that the regression equation generation unit classifies past total demand data based on the frequency of total demand for each day.

7. The load control device according to claim 4, characterized in that the regression equation creation unit creates multiple regression equations by fitting multiple regression equations to past total demand data so that the regression error of each regression equation follows a normal distribution, thereby grouping past total demand data and creating multiple regression equations.

8. The load control device according to claim 1, characterized in that the air conditioning equipment demand estimation unit calculates a reference value for the demand of the air conditioning equipment and a reference value for the total demand of other load equipment other than the air conditioning equipment based on the regression equation, calculates the regression error of the regression equation, and estimates the current demand for the air conditioning equipment by apportioning the calculated regression error.

9. The load control device according to claim 8, characterized in that the air conditioning equipment demand estimation unit apportions the regression error by the ratio of the reference value of the demand for the air conditioning equipment to the reference value of the total demand for the other load equipment.

10. The load control device according to claim 8, characterized in that the air conditioning equipment demand estimation unit apportions the regression error using the ratio of the regression error between the intermediate period when the air conditioning equipment is not in operation and the period when the air conditioning equipment is in operation in historical total demand data.

11. The regression equation creation step involves creating a regression equation that expresses the demand for air conditioning equipment as temperature-sensitive demand based on historical aggregate demand data, and A step to estimate the demand for air conditioning equipment involves obtaining total demand data and estimating the current demand for air conditioning equipment using the regression equation created in the regression equation creation step, A demand control step that calculates a predicted total demand at a predetermined time, and controls the air conditioning equipment based on the predicted value and the current demand for the air conditioning equipment estimated in the air conditioning equipment demand estimation step, so that the total demand at the predetermined time falls within a predetermined demand target value. A load control program characterized by having a computer execute it.

12. The load control device The process involves creating a regression equation that expresses the demand for air conditioning equipment as temperature-sensitive demand based on historical aggregate demand data, and A process for estimating the current demand for air conditioning equipment involves acquiring total demand data and estimating the current demand for air conditioning equipment using the regression equation created in the regression equation creation process, A demand control process that calculates a predicted total demand at a predetermined time, and controls the air conditioning equipment based on the predicted value and the current demand for the air conditioning equipment estimated by the air conditioning equipment demand estimation process, so that the total demand at the predetermined time falls within a predetermined demand target value. A load control method characterized by performing the following:

Images