WO2012114974A1

WO2012114974A1 - Apparatus for forecasting electric power load of industrial park, system and method

Info

Publication number: WO2012114974A1
Application number: PCT/JP2012/053670
Authority: WO
Inventors: 鈴木　勝幸; 正教神永; 文乃田中
Original assignee: 株式会社日立製作所
Priority date: 2011-02-23
Filing date: 2012-02-16
Publication date: 2012-08-30
Also published as: JP2012175847A; CN103380557A

Abstract

In the present invention, a scheme is constructed for forecasting an electric power load taking into consideration the operation patterns of each factory, and a means is provided in which it is possible to make forecasting calculations even if there are operation patterns for which there are no past operation records. An apparatus for forecasting the electric power load of an industrial park is designed to forecast the amount of electric power required on a specific day in the future by an industrial park composed of one or a plurality of plants. In plants there is production equipment contributing to production and non-production equipment not contributing to production. Regarding the production equipment, a high-quality electric power forecasting unit is provided for forecasting the progress of the amount of electric power at each time of a specific day using a production plan for a specific day in the future and a device and the usage time of the device used in the production plan. Regarding the non-production equipment, a low-quality electric power forecasting unit is provided for accumulating the progress of the amount of electric power of the non-production equipment in the past and forecasting the progress of the amount of electric power at each time of a specific day in the future. Regarding the progress of the amount of electric power at each time of a specific day calculated by each unit, calculating the sum of each unit for the same time of day gives the amount of electric power required by the industrial park on the specific day in the future.

Description

Industrial estate power load prediction apparatus, system and method

The present invention relates to a power load predicting apparatus, system and method for an industrial park, and more particularly to a power load predicting apparatus for an industrial park which is necessary for realizing system stabilization and economical operation of private power generation facilities for industrial parks. , Systems and methods.

In recent years, research and development of smart grids and microgrids that incorporate information and communication technologies has been active for the purpose of realizing low-carbon and economical power operation. On the other hand, in some countries and regions, commercial grid power is fragile and unstable, increasing the number of cases that interfere not only with daily life but also with factory production activities. When moving production bases from Japan to overseas, how to secure high-quality power is a big problem. In this regard, each company is making efforts as shown in Non-Patent Document 1 and Non-Patent Document 2.

For example, in Non-Patent Document 1, a diesel generator is installed in each factory and used for power supply in order to ensure high quality power necessary for factory production. When the production load is high, cooperation with the commercial system is also performed, but basically diesel generator cooperation. When factory production is stopped due to holidays, diesel generators are also stopped, but there are examples of power interchange to other factories using recent microgrid technology.

In this case, it is necessary to operate the diesel generator economically. For example, it is possible to solve the diesel generator optimum operation planning problem with the objective function of minimizing the consumption of diesel fuel.

In realizing the optimum operation plan of the private power generation facility, Non-Patent Document 1 shows the diesel generator cooperative control of the private power generation facility based on the prediction of the power load. Further, in Non-Patent Document 2, the operation plan for the next day is calculated using the actual power data up to the previous day.

In these already proposed power load prediction methods, past operation results are used to extract data estimated to be similar to predict rough load changes. After that, when the measurement value for the day is available, correction is performed in consideration of the prediction error.

As described above, the conventional power load prediction method is based on the assumption that past operation data of the factory is used.

However, when microgrids such as industrial parks are targeted, there are few cases where the operation result data of each factory is centrally managed. In addition, the tendency of power load changes often varies. That is, there are various cases such as an assembly line, a case where an intermittent load change occurs such as carrier power, and a case where a power load continues for a long time like a chemical process. In the above-mentioned document, a solution for such a problem is not clearly described.

From the above, it is considered that there are two issues to be solved. The first issue is the realization of a prediction method based on the change in power load at each factory. Even if a large number of operation result data are prepared, the prediction result does not match if an operation pattern that does not exist in the past is executed.

The second issue is the realization of a method for estimating the power load necessary for planning an operation plan for an economical private power generation facility in order to ensure high-quality power. The system instability should not be caused as a result of reducing the number of in-house power generation facilities in operation due to economic considerations.

In order to achieve the above object, the present invention aims to provide a power load prediction method that takes into account the operation pattern of each factory, and to provide means capable of predictive calculation even when there is an operation pattern that has not been operated in the past. And

In order to achieve the above object, in the present invention, in the power load predicting device of an industrial park for predicting the amount of power required in an industrial park composed of one or a plurality of establishments for a specific date in the future. Among the production facilities that are involved in production at the office and non-production facilities that are not involved in production, the production facilities are identified using the production plan on a specific date in the future, the equipment used in this production plan, and the time of use. For the high-quality power forecasting unit that predicts the daily power consumption for each time, and for non-production facilities, it accumulates the past power consumption transitions of the non-production facilities, and based on past results, Equipped with a low-quality power forecasting unit that predicts the amount of electricity for each time, and for the amount of electricity obtained for each time, finds the sum for each time and is required in the industrial park on a specific date in the future The amount of power.

The high-quality power prediction unit includes a production facility power characteristic database that stores the equipment used here and the rated power amount for each production plan, and the low-quality power prediction unit stores the non-production facility in the past. It has a power load record database that stores changes in power consumption.

In addition, the industrial park is composed of a plurality of business establishments, and the high-quality power prediction unit and the low-quality power prediction unit are provided for each business establishment in the industrial park and perform prediction of electric energy.

In addition, the industrial park is composed of a plurality of business establishments, and the high-quality power forecasting unit is provided for each business establishment in the industrial park to perform power forecasting, and the low-quality power forecasting unit is installed in the industrial park. Predicts the amount of power that is commonly provided at multiple offices.

In order to achieve the above object, in the present invention, an industrial park composed of one or a plurality of business establishments connected by a power transmission line, a private power generation power source that supplies power to the business establishment, In a power load prediction system for an industrial park that is composed of a power load prediction device for predicting a specific date in the future, the power load prediction device is a non-production device that is not involved in production and a production facility that is involved in production at an office. Among the facilities, a high-quality power prediction unit that predicts changes in the amount of power for each specific day by using the production plan for a specific date in the future, the equipment used in this production plan, and its usage time For non-production facilities, a low-quality power forecasting unit that accumulates the transition of power consumption of non-production facilities in the past and predicts the transition of power consumption for each specific day in the future based on past results. For example, the transition of each time power amount calculated respectively, and the amount of power needed in the industrial in certain date in the future to seek the sum of each same time.

In addition, the power generation amount of the private power generation power source is controlled on a specific date in the future based on the transition of the power amount predicted by the power load prediction device.

In order to achieve the above object, in the present invention, in an industrial park power load prediction method for predicting the amount of power required in an industrial park composed of one or more offices for a specific date in the future. , Classify the electric power facilities of the office into production facilities that are involved in production and non-production facilities that are not involved in production, and for production facilities, the production plan on a specific date in the future, the equipment used in this production plan, and the time of use Is used to predict changes in the amount of electricity on a specific day for each time, and for non-production facilities, the past changes in the amount of non-production equipment in the past are accumulated. Predict the transition of each time.

In addition, regarding the transition of the obtained electric energy for each time, the sum for each hour is obtained to obtain the electric energy required in the industrial park on a specific date in the future.

According to the power load predicting apparatus, system and method of the industrial park of the present invention, it is possible to estimate a power load change even when there is no past operation pattern, and stable operation of private power generation such as a diesel generator is possible. In addition, it is possible to estimate the power load peak with high accuracy by adding past applicable data such as office air conditioning and lighting to the estimation calculation, and as a result, it is possible to stabilize the diesel generator. It becomes possible to execute the operation plan.
Other objects, features and advantages of the present invention will become apparent from the following description of embodiments of the present invention with reference to the accompanying drawings.

The figure which shows the electric power load prediction system of this invention. The figure which shows the structural example of the industrial complex to which this invention is applied. The figure which shows the example which operates during the day as an example of a production plan. The figure which shows an example of the production facility electric power characteristic database in the daytime operation example. The figure which showed the relationship between manufacturing equipment operating time and electric power transition in the daytime operation example. The figure which shows the view of low quality electric power prediction. The figure which shows the electric power load prediction result of the high quality and the low quality in the daytime operation example. The figure which shows the example which operates continuously as an example of a production plan. The figure which shows an example of the production facility electric power characteristic database in the example of continuous operation. The figure which showed the relationship between the manufacture apparatus operating time and power transition in the example of continuous operation. The figure which shows the electric power load prediction result of the high quality and the low quality in the continuous operation example. The figure which shows the example which mainly operates at night as an example of a production plan. The figure which shows an example of the production facility electric power characteristic database in the example of night operation. The figure which showed the relationship between manufacturing equipment operating time and electric power transition in the example of nighttime operation. The figure which shows the electric power load prediction result of the high quality and the low quality in the daytime operation example.

An embodiment of a power load prediction apparatus and system according to the present invention will be described below with reference to the drawings.
While the above description has been made with reference to exemplary embodiments, it will be apparent to those skilled in the art that the invention is not limited thereto and that various changes and modifications can be made within the spirit of the invention and the scope of the appended claims.

First, the configuration of an industrial park 100 to which the present invention is applied will be described with reference to FIG. In this industrial park 100, a plurality of offices 10 (three

units

10A, 10B, and 10C) and a power source 20 are installed. The plurality of business establishments 10 are connected to the power source 20 via the power transmission line 40 and are supplied with power. Although the power supply 20 may be installed outside the site of the industrial park 100, it is often a private power generation facility that supplies power exclusively to establishments in the industrial park, and is composed of a plurality of diesel generators with good responsiveness. Is good. In addition, the plurality of business establishments 10 can be connected to an external commercial power supply 30 and receive power as necessary as a countermeasure when power supply from the power supply 20 is insufficient.

Each facility 10 has various facilities as an electric power load, and these facilities are roughly classified into a production facility L1 and a non-production facility L2. The production facility L1 is equipment and facilities that are directly operated and stopped according to the production plan and operation plan of the factory. The non-production facility L2 is equipment in the office, lighting equipment in the office and factory, and production. Utilities for facilities are included. However, utilities for production facilities may be classified into production facilities L1.

The power supplied to the production facility is referred to as production facility power P1, and the power supplied to the non-production facility is referred to as non-production facility power P2. In general, the production facility power P1 is sometimes referred to as high-quality power. Accordingly, in the present invention, the non-production facility power P2 is referred to as low-quality power.

Although it may be inside or outside the industrial park 100 in FIG. 2, the power load prediction device 50 of the present invention is installed, and the power demand prediction of the industrial park on a specific date in the future (hereinafter referred to as X day) is executed.

For this reason, the past load value of the power P2 for non-production facilities measured at each business office 10 is captured and accumulated in the power load prediction device 50. In addition, a production plan S for a specific date in the future (X days) of each office is given. As a result, the power load prediction device 50 obtains a power load prediction value Y (X) for a specific date in the future. The predicted power load value Y (X) may be given as reference information for the manager 60 who manages the private power generation power source 20, or may be a direct control signal for the diesel generator DEG that constitutes the power source 20.

The power load prediction apparatus 50 of the present invention is configured as shown in FIG. In this figure, 50A is a power load predicting apparatus (hereinafter referred to as a partial power load predicting apparatus) for the office 10A, and this partial power load predicting apparatus is provided for each office in the industrial park. 50B and 50C are partial power load prediction apparatuses for the offices 10B and 10C, respectively, and have the same configuration as 50A. Although these partial power load prediction apparatuses are different in data to be held, but the basic processing contents are the same, the following description will be made with an example of the partial power load prediction apparatus 50A for the office 10A. In addition, from the power load prediction apparatus 50 of FIG. 1, the power load prediction value Y (X) of the future specific day of this industrial park is finally obtained. The predicted power load value Y (X) is a time-series total power load amount of the industrial park.

The partial power load prediction apparatus 50A includes a power load prediction unit 1 for production facilities (hereinafter referred to as a high quality power load prediction unit) and a power load prediction unit 2 for non-production facilities (hereinafter referred to as a low quality power load prediction unit). ), And the respective predicted values YA1 (X) and YA2 (X) are added together to obtain the partial power load predicted value YA (X) of the office 10A. In the partial power load prediction apparatus 50A of the present invention, the power load prediction value YA1 (X) for production facilities and the power load prediction value YA2 (X) for non-production facilities are derived by the best method suitable for each. To do.

For this purpose, the high-quality power load prediction unit 1 determines the date and time of the future specific date X in the prediction date and time setting unit 13, and sets the production plan SA on the date and time of the specific date X obtained from the office 10A as the production plan setting unit. 16 is held. In addition, the amount of power required for each production facility is stored in the production facility power characteristic database 11.

An example of the production plan SA that operates during the day will be described with reference to FIG. This office 10A is a factory that manufactures resin compounds, for example, and basically operates from 8:00 to 18:00 on weekdays and stops at night and on weekends. In the resin compound production, the material kneading process Pr1, the granulation process Pr2, the cooling process Pr3, and the cutter / metering process Pr4 are executed. Each process Pr and its execution time are given in advance as a production plan. In the example of FIG. 3, an example is shown in which all the processes Pr are executed simultaneously from 8:00 to 18:00.

The production facility power characteristic database 11 stores the production facility for executing each process and its power load capacity (rated power consumption), and a specific example is shown in FIG. In this figure, for example, for the material kneading process Pr1, it is necessary to operate power devices having device numbers M11 to M14 as production facilities, and the power load capacity of these power devices is the device number M11. 200 kW, device number M12 is 1000 kW, and the like. In this example, the power consumption (1000 kW) of the extruder (equipment number M12) used in the material kneading process Pr1 is large, which is considered to be a main factor of power fluctuation. Similarly, the production facility power characteristic database 11 stores the relationship between the equipment to be used and the amount of power for the other processes Pr2-Pr4.

The production facility power characteristic data includes the rated power value based on the facility capacity unique to the manufacturing facility and the time variation of the operating load. Specifically, the operation time from the start to the end of the operation of the equipment and the corresponding power load change. In the present invention, since the purpose is to confirm the peak value of the power load, the power characteristic in the operating time range may be a constant value equal to the equipment capacity.

FIG. 5 shows the relationship between specific manufacturing equipment operating time and power transition in the material kneading process Pr1. The manufacturing equipment operating time indicates the manufacturing equipment operation plan, that is, the manufacturing equipment operation stop schedule associated with the production plan. The manufacturing equipment operating time may be associated with the process and stored in the production plan 16, or if the process is determined, the equipment to be used and its operating time are uniquely determined. It may be stored in the facility power characteristic 11.

In the example of FIG. 5, as shown in the middle part of FIG. 5, the device M11 operates during the entire period of the process Pr1, and the device M12 that is the largest power load operates from 9:00 to 11:00. The device M13 operates from 11:00 to 16:00, and the operating device M14 operates from 16:00 to 18:00.

The operation scheduled facility search unit 12 in FIG. 1 obtains the process name to be executed at each time of X days from the production plan setting unit 16, and uses (operates) the production facility to be used with the obtained process name and its power load capacity. Is obtained from the production facility power characteristic database 11.

The time / power combining unit 14 continuously obtains the power load capacity at each time on the X day, and calculates the transition of the power load capacity at 24 hours on the X day. That is, a device that is operating at each time is extracted, and the power load amount is added to obtain a power transition for 24 hours on the X day.

This example is shown as a power load at the bottom of FIG. In the case of this material kneading process Pr1, the apparatus M11 is used from 8:00 to 9:00, and the power load at this time is 200 kW. The devices M11 and M12 are used from 9:00 to 11:00, and the power load at this time is 1200 kW. The transition and detailed explanation are omitted, but the transition is made to 400 kW and 300 kW, and the operation of this day is stopped.

FIG. 5 shows an example of obtaining the power transition for 24 hours on the X day in the case of the material kneading process Pr1, but the power transition for 24 hours on the X day is similarly calculated for other processes. After that, the power transition for 24 hours in total X days of each process is obtained. Although the total power amount transition is not illustrated, in the following description, the power load amount transition pattern shown in the lower part of FIG. 5 is the predicted value of the high-quality power load prediction unit 1 in order to simplify the description. A description will be given assuming that YA1 (X). The transition of the power load capacity in 24 hours on the X day thus obtained is output from the high quality power predicted value output unit 15 as the predicted value YA1 (X).

Next, in the low quality power load prediction unit 2 of FIG. 1, the prediction date setting unit 22 determines the future specific date X, includes the operation calendar 24, and the non-production facility power P2 measured in the office 10A. Past performance values are taken in and stored in the power load performance database 21.

Here, low-quality power (power load not connected to manufacturing equipment) refers to equipment that is not directly related to manufacturing, such as office lighting and air-conditioning power, as described above. This is called low quality power because it does not affect product quality. In many cases, it may be considered that there are characteristics that are influenced by weather factors such as temperature and humidity in addition to calendar information such as date and time.

Figure 6 explains the concept of low-quality power prediction. The daily power load record stored in the power load record database 21 indicates, for example, fluctuations in the power load within 24 hours as shown in FIG. Since this power load record is a past record value of the non-production facility power P2, the power load such as lighting and cooling / heating is mainly used, and it increases in the daytime when people are in the office as a whole, and decreases at nighttime. Furthermore, there are different tendencies between the operating days and the non-operating days, and there are also day and seasonal fluctuations. In addition, the weather of the day will also have an effect. The power load record database 21 stores past record values of non-production facility power P2 measured under these various conditions.

The upper part of Fig. 6 shows the standard pattern of power load for office lighting and air conditioning in the factory. As a characteristic of the factory, it is assumed that the low-quality power is equal to or less than the power load of the manufacturing process. In the standard pattern in the upper part of FIG. 6, the operation time zone is a maximum of 500 kW load, and the other time zone is approximately 200 kW load. In the low quality power prediction, in addition to the standard pattern, the previous day actual value or the power load actual on the same day of the previous week is referred to.

The previous day's actual values in the middle of Fig. 6 have data with a maximum 500 kW power load, as in the upper standard pattern, but with sudden changes in the morning load. The result of predicting a low-quality power load with reference to these standard patterns and the previous day actual values is shown in the lower part of FIG. There are various methods for obtaining the predicted value from the standard pattern and the previous day actual value, but generally a weighted weighted average value is obtained for each time and used as the predicted value. As a result, the result of the lower solid line can be obtained.

The similar load search unit 25 of the low quality power load prediction unit 2 obtains the past actual value of the power P2 for non-production equipment that matches the condition of the specific date X in the future from the power load result database 21. Note that the past performance value obtained here may be obtained as an average value of several similar pieces of data that match the conditions, as well as extracting one piece of data.

The similar load search of the low-quality power load prediction unit 2 may be executed based on the above-mentioned several ideas, and a typical example is introduced in FIG. In this case, the standard pattern in FIG. 6 is obtained from the past results as an average value of several pieces of similar data that meet the conditions. In this pattern, the amount of power in the daytime when people are present increases and decreases relatively slowly, and peaks around noon. On the other hand, in the data of the most recent previous day (in FIG. 6), there is a noticeable tendency that the amount of power increases rapidly from before the start of work and decreases rapidly with retirement. Taking these two patterns into consideration, the final power forecast reflects the tendency of the previous day's performance at the time of going to and from work, and the solid line at the bottom of Fig. 6 takes advantage of the characteristics of the standard pattern in which the amount of power peaks around noon. Get power predictions for The transition of the power load capacity in 24 hours on the X day obtained in this way is output from the low quality power predicted value output unit 23 as the predicted value YA2 (X).

In addition, although the low quality power load prediction unit 2 in the power load prediction device 50A in FIG. 1 has been described to predict the low quality power load of the office 10A, this predicts the low quality power load for each office. Instead, the system can be configured more simply than the case where the low quality power load prediction is performed for each business establishment by collectively predicting the low quality power load as the entire industrial park. That is, the low quality power load prediction unit 2 is installed only in 50A to collectively predict the low quality power load as the entire industrial park, and the low quality power load prediction unit 2 is not provided in the

other prediction units

50B and 50C. It is good to have a configuration.

FIG. 7 shows the high-quality power transition YA1 (X) and the low-quality power YA2 (when the establishment 10A in FIG. 2 operates on a specific date X in the future based on the production plan (processes Pr1 to Pr4 are operated in the daytime). This is a result of predicting X) along the time direction. However, in this figure, as described above for convenience of explanation, the pattern of the prediction result of process Pr1 in FIG. 5 (bottom of FIG. 5) is the total high-quality power prediction result including other processes. The same pattern is shown in FIG. In addition, in order to actually obtain the final integrated power load predicted value YA (X) of the factory concerned, high quality power is obtained for each of the other processes P2, P3, and P4 in the same manner as in FIG. 5, and cumulatively added. You can get it.

Referring back to FIG. 1, the total power load predicted value YA (X) of the factory is obtained in the same manner for the other offices 10B and 10C in the industrial park. This method of obtaining is basically easy to understand from the above description, so only an example will be described here.

The office 10B is assumed to be an assembly factory such as an automobile or an air conditioner as a case of full production for 24 hours. FIG. 8 shows a production plan in the case of full production for 24 hours. Here, the processes Pr5 to Pr8 are continuously operated.

FIG. 9 is a diagram showing 11 examples of the production facility power characteristic database in the continuous operation example, and particularly shows the relationship between the manufacturing equipment and the rated power consumption in the process Pr5. In this table, devices M51, M52, and M53 and their rated power consumption W51, W52, and W53 are described for the process Pr5. The other processes Pr6 to Pr8 are similarly set.

FIG. 10 shows the production equipment operation stop schedule in the process Pr5 in the continuous operation example in the middle. In the process Pr5, the devices M51 and M52 are in a full operation state day and night, and the M53 operates from 2 o'clock to 10 o'clock. The corresponding power load predicted value is shown in the lower part of FIG. 10, and a power load of 950 kW or 1000 kW is obtained as the predicted value through 24 hours. From 21:00 to 23:00, the power load drops to 930 kW, which is due to a decrease in the nighttime production load. In FIG. 9, when the partial load operation data of the manufacturing equipment is not described, it is estimated to be 950 kW. There is no problem.

FIG. 11 shows an example in which the high quality power prediction value YB1 (X) and the low quality power prediction value YB2 (X) as prediction results are written in the time direction. In the illustrated explanation, the prediction result waveform of the process Pr5 in FIG. 10 is marked as a final high-quality power prediction value YB1 (X) that also includes calculation results in other processes.

Moreover, since the low quality prediction value changes regardless of the operation schedule of the manufacturing equipment as in the case of the daytime operation in the previous period, the prediction value obtained for the industrial park collectively in FIG. 6 was used here.

The office 10C in FIG. 2 is an example of full operation day and night, but as shown in FIG. 12, in addition to the 24-hour operation process Pr9, it is a night-dominated type that includes night operation processes Pr10, Pr11, and Pr12. It takes the form of operation. In FIG. 12, since the next day after 23:00 is the next day's data, the process Pr10 is shown from 0 o'clock to 6 o'clock, but the actual operation is a process that runs at 20 o'clock and stops at 6 o'clock the next day. .

FIG. 13 is a data table of rated power consumption corresponding to the manufacturing equipment number of the process 10. Here, for the device M102, the equipment activation time is shown in parentheses. The device M102 has a large rated power capacity, such as an induction heating furnace, and requires a long time for startup, and it is necessary to add a power consumption increase pattern for each time.

FIG. 14 is a diagram showing the manufacturing equipment operation stop schedule in the middle in the process Pr10. In this schedule, the devices M101 and M102 are activated at 20:00 at night, the M102 is deactivated at 4am the following day, the device M103 is activated instead, and the device is deactivated at 6am. The predicted power load value corresponding to this is shown in the lower part of FIG. Since the startup time of 1.5 hours is required for M102 among the facilities operating at 20:00 at night, a prediction result of reaching the rated power consumption at time 21:30 is obtained. After that, in the next day predicted value calculation, a result that the power consumption predicted value decreases from 1100 kW to 250 kW at 4 o'clock is output.

FIG. 15 is an example in the process Pr10 showing the high-quality power predicted value in the middle and the low-quality power predicted value in the lower level in the time direction. The low quality power prediction value is obtained in the same manner as in FIG. The predicted power load value for the entire factory is obtained by cumulatively adding the process Pr9 operating for 24 hours and the other processes Pr11, Pr12, and Pr13.

As described above, the power load prediction in the case where the factory is operated during the daytime, full production for 24 hours, and night operation has been shown. However, the factory holiday can be dealt with by using the low quality power prediction value shown in FIG. . However, it is necessary to prepare a standard pattern for a factory holiday.

As described above in detail, in the power load prediction according to the present invention, the high quality power load prediction and the low quality power load prediction are executed, and the prediction values from these are added to obtain the final power load prediction value. . The former is a power load prediction based on a future production plan, and the latter is a power load prediction based on past performance.

In FIG. 1, the high quality power prediction value YA1 (X) and the low quality power prediction value YA2 (X) of the office 10A are added at the same time in the adding means 3, and as a result, the power load prediction value 4 (YA ( X)).

In the present invention, the predicted load values YB (X) and YC (X) of the

other factories

50B and 50C are received as notifications, and added by the adding means 5, whereby the total power load predicted total value Y (X )

As described above, the prediction result is automatically reflected in the direct control of the plurality of diesel generators DEG in the private power generation power source 20, or in the manual setting based on the judgment of the administrator.

The high-quality power obtained in this way is directly connected to the diesel generator group DEG, which is a private power generation facility, and even when the external system power supply is unstable, stable high-quality power can be obtained. Is possible. On the other hand, low-quality power is allowed to change in voltage and the like, and is configured to be directly connected to an external power supply. However, the diesel generator and the external system can be connected, and low-quality power can be supplied by the diesel generator.

As described above, the power load prediction method of the present invention enables accurate load prediction for a power load that requires high quality such as voltage stability based on a production facility operation plan. Moreover, since the load prediction value is obtained separately from the prediction of low quality power such as miscellaneous power, it is possible to consider a load change with periodicity.

DESCRIPTION OF SYMBOLS 1 Power load prediction part for production facilities 2 Power load prediction part for non-production facilities 5 Power

load prediction devices

10A, 10B, 10C Multiple establishments 11 Production facility power characteristic database 13 Prediction date setting part 14 Time / power Combining unit 16 Production plan setting unit 20 Power supply 21 Power load record database 24 Operation calendar 25 Similar load search unit 30 Commercial power supply 40 Transmission line 50A Power

load prediction device

50B, 50C For business sites 10B, 10C Partial power load prediction device 100 Industrial park L1 Production facilities L2 Non-production facilities P1 Power for production facilities P2 Power for non-production facilities DEG Diesel generator M Equipment

Claims

In an industrial park power load forecasting device for forecasting the amount of power required in an industrial park composed of one or more establishments for a specific date in the future,
Of the production facilities involved in production at the office and non-production facilities not involved in production, for the production facilities, the production plan for a specific date in the future, the equipment used in this production plan, and its use time , A high-quality power prediction unit that predicts the transition of the power amount of the specific day for each time, and
For the non-production facility, the transition of the power amount of the non-production facility in the past is accumulated, and a low-quality power prediction unit that predicts the transition of the power amount for each specific time in the future based on past results, respectively, A power load predicting device for an industrial park characterized in that, with respect to the transition of the power amount obtained in step 1, the sum at each time is obtained to obtain the amount of power required in the industrial park on a specific date in the future.
In the electric load prediction apparatus of the industrial park of Claim 1,
The high-quality power prediction unit includes a production facility power characteristic database for storing the equipment used here and the rated power amount for each production plan, and the low-quality power prediction unit A power load prediction device for an industrial park, comprising a power load record database for accumulating changes in power consumption.
In the industrial area electric power load prediction device according to claim 1, wherein the industrial area is composed of a plurality of establishments.
The high-quality electric power prediction unit and the low-quality electric power prediction unit are provided for each business establishment in the industrial park and execute the prediction of the amount of electric power.
In the industrial area electric power load prediction device according to claim 1, wherein the industrial area is composed of a plurality of establishments.
The high-quality power prediction unit is provided for each business establishment in the industrial park, and performs prediction of electric energy,
The low-quality power prediction unit is provided in common to a plurality of business establishments in an industrial park and executes a power amount prediction.
Industrial parks composed of one or more establishments connected by power transmission lines, private power generation power to supply the establishments, and electricity to predict the amount of power required in the industrial estate for a specific date in the future In an industrial park power load prediction system consisting of a load prediction device,
The power load prediction device includes a production plan for a specific date in the future, and a device used in the production plan for the production facility among production facilities involved in production at the office and non-production facilities not involved in production. And the high-quality power prediction unit that predicts the transition of the power amount of the specific day for each time using the time of use, and the non-production facility, accumulates the transition of the power amount of the non-production facility in the past, Equipped with a low-quality power forecasting unit that predicts the transition of the amount of power for a specific day in the future based on past results. A power load prediction system for an industrial park, characterized in that the amount of power required in the industrial park on a specific date of the industrial park.
In the industrial area electric power load prediction system according to claim 5,
The high-quality power prediction unit includes a production facility power characteristic database for storing the equipment used here and the rated power amount for each production plan, and the low-quality power prediction unit A power load forecasting system for an industrial park, characterized by having a power load record database for accumulating changes in power consumption.
In the industrial estate power load prediction system according to claim 5, wherein the industrial estate is composed of a plurality of establishments.
The high-quality power prediction unit and the low-quality power prediction unit are provided for each business office in the industrial park, and execute a power amount prediction.
In the industrial estate power load prediction system according to claim 5, wherein the industrial estate is composed of a plurality of establishments.
The high-quality power prediction unit is provided for each business establishment in the industrial park, and performs prediction of electric energy,
The low-quality power prediction unit is provided in common to a plurality of business establishments in an industrial park and executes a power amount prediction.
In the industrial area electric power load prediction system according to claim 5,
A power load prediction system for an industrial park, wherein the power generation amount of the private power generation power source is controlled on the specific date in the future based on the transition of the power amount predicted by the power load prediction device.
In the method of predicting the power load of an industrial park for predicting the amount of power required in an industrial park composed of one or more establishments for a specific date in the future,
The electric power facilities of the business office are classified into production facilities that are involved in production and non-production facilities that are not involved in production. Using the time of use, predict the transition of the amount of power on the specific day for each time, accumulate the transition of the power amount of the non-production facility in the past for the non-production facility, and identify the future based on the past performance A method for predicting the power load of an industrial park, wherein the transition of the daily power consumption at each time is predicted.
In the electric load prediction method of the industrial estate of Claim 10,
A method for predicting an electric load of an industrial park, characterized in that, for each transition of electric power obtained for each time, a sum for each time is obtained and used as an electric energy required for the industrial park on a specific day in the future.
In the industrial park electric power load prediction apparatus according to claim 2, wherein the industrial park is composed of a plurality of establishments.
The high-quality electric power prediction unit and the low-quality electric power prediction unit are provided for each business establishment in the industrial park and execute the prediction of the amount of electric power.
In the industrial park electric power load prediction apparatus according to claim 2, wherein the industrial park is composed of a plurality of establishments.
The high-quality power prediction unit is provided for each business establishment in the industrial park, and performs prediction of electric energy,
The low-quality power prediction unit is provided in common to a plurality of business establishments in an industrial park and executes a power amount prediction.
In the industrial estate power load prediction system according to claim 6, wherein the industrial estate is composed of a plurality of establishments.
The high-quality power prediction unit and the low-quality power prediction unit are provided for each business office in the industrial park, and execute a power amount prediction.
In the industrial estate power load prediction system according to claim 6, wherein the industrial estate is composed of a plurality of establishments.
The high-quality power prediction unit is provided for each business establishment in the industrial park, and performs prediction of electric energy,
The low-quality power prediction unit is provided in common to a plurality of business establishments in an industrial park and executes a power amount prediction.