JP2000333367A - Thermoelectric consumption predicting system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a thermoelectric consumption predicting system for optimizing energy efficiency of an energy plant, without making the system complex in configuration. SOLUTION: This thermoelectric consumption predicting system determines an operation schedule in an energy plant for supplying either of heat and power to at least one supply target of heat and power and is provided with a communication device 1 for reading weather-related data from the outside and an operation device 2 for outputting an least one of a heat consumption prediction value and a power consumption prediction value, based on weather data read from a communication device 1.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an operation control of an energy plant for supplying heat and electricity to a building or a specific area, and more particularly, to predicting a demand for heat and electricity used for determining an operation schedule of an energy plant. It relates to a thermoelectric demand forecast system.

[0002]

2. Description of the Related Art Heretofore, the operation efficiency of an energy plant that supplies heat and electricity to an area largely depends on its operation schedule. In particular, when the energy plant has a heat storage tank, an operation is performed in which a large amount of heat is generated using inexpensive nighttime electric power and stored in the heat storage tank, and the heat stored in the heat storage tank is used for cooling and heating in the daytime.

[0003] In an energy plant using such a heat storage tank, cold energy is stored based on the demand forecast for cooling and heating performed on the previous day. Therefore, the effect of the heat storage operation becomes insufficient.

To this end, various prediction methods have been proposed and put into practical use. Among the many methods that have been proposed, a typical method is to store past demand patterns as actual results, and to record past results for days with similar conditions (mainly days of the week are used). And a method using case-based reasoning to extract similarity with past results.

Another typical method is a method using a weather forecast. This method is a method of calculating a predicted value by using weather forecast data, particularly, a maximum temperature and a minimum temperature, and utilizing characteristics of demand, for example, features such as a commercial building and a house.

[0006] Many of these techniques are compromised. This is a method of combining past demand actual patterns and weather forecast data. Some systems adjust this combination by learning.

[0007]

The demand for heat and power depends on the activity of the person peculiar to the day and the weather. The systematic part, such as day-of-the-week dependence, can be predicted from past patterns, while the random part is difficult to predict using a demand forecast system in general. is there.

[0008] On the other hand, a portion which is affected by the weather can be accurately predicted if the prediction of the weather is accurate.

In the prior art, in the case of a method of predicting based only on past demand pattern results, it is possible to predict a systematic portion by capturing results such as day-of-week dependence of human activities. However, it is not possible to take into account the influence of weather, which has a large effect on demand. Although the actual weather can be entered, there is no mechanism to accurately input the necessary tomorrow's weather.

[0010] On the other hand, in the method in which the weather forecast is added to the actual result, the best prediction is obtained within the range of the given information.

However, in the conventional method, since the input is left to the operator of the thermoelectric demand prediction system, it is troublesome to actually input the information, and a highly accurate input cannot be expected. In particular, it has been difficult for general weather forecasts to input a predicted value of sunshine which has a large effect on heat and power demand. Also, it has been difficult to trace and input a weather forecast that changes every moment. Due to such difficulties, it has been difficult to increase the accuracy of the thermoelectric demand forecasting system to a sufficient level, and thus it has been difficult to optimize the energy efficiency of the district heat source plant and the cogeneration plant.

Further, in the conventional prediction system, as described above, the power demand prediction and the heat demand prediction have been developed as completely separate systems while having many common elements. As a result, in a cogeneration plant or the like, there is a problem in that it is necessary to prepare two overlapping means, resulting in a complicated system configuration.

An object of the present invention is to provide a thermoelectric demand forecasting system capable of optimizing the energy efficiency of an energy plant without complicating the system configuration.

[0014]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides a system for predicting the demand for heat and power in an energy plant that supplies at least one of heat and power to at least one of heat and power. A communication device that takes in weather-related data from the outside,
A computing device that executes at least one of heat demand prediction and power demand prediction based on the weather data captured by the communication device, and outputs the result.

[0015] With this configuration, the communication device fetches the national data of AMeDAS and visible images of satellites, which are released as weather-related data, every moment. In addition, the temperature included in the information taken from the communication device by the arithmetic device,
By extracting trend information of sunshine and cloud nationwide data, grasping the changing trends of temperature and sunshine, based on the changing trends,
The temperature and sunshine information at the time to be predicted are predicted, and the day of the information is input to functions such as weather condition dependence and day dependence determined from past cases to calculate heat demand and power importance. Furthermore, with the communication device and the arithmetic device,
Correct and forecast heat demand and electricity demand.

[0016]

Embodiments of the present invention will be described with reference to the drawings.

An embodiment of a thermoelectric demand forecasting system according to the present invention, as shown in FIG. 1, uses a communication device 1 connected to a telephone line and a sunshine and temperature based on weather information obtained by the communication device 1. An image processing device 2 for creating prediction data of the above, a function operation device 3 for predicting heat demand and power demand based on the data created by the image processing device 2, a heat and power demand for sunshine, temperature and date A demand learning device 4 for obtaining a daily change curve of the actually measured value by a regression curve.

Here, the communication device 1 can be easily realized by connecting a modem installed in a personal computer to a telephone line and periodically receiving weather data. The communication device 1 suppresses communication costs by cutting up communication after obtaining information by dialing up at regular time intervals. As a result, it is possible to always receive the latest data such as the temperature and sunshine of AMeDAS installed nationwide on the Internet as shown in FIG. Further, by changing the settings in the communication device 1, the latest weather forecast data of the Meteorological Agency can be obtained.

Next, the image processing device 2 stores the temperature and sunshine data of the whole country as shown in FIG. 2 at appropriate intervals.
Classification is performed on the sunshine data as shown in FIG. In order to eliminate the influence of special terrain and the like, the AMeDAS measurement points (illuminance and temperature) are approximated by least squares approximation as an n-th surface on a two-dimensional surface of the X and Y axes.

In general, it is preferable that n be simplified to about the second order.

Z = aX ² + bY ² + cXY + dX + eY + f (1) As the contour line, the class can be easily obtained. This curved surface is sequentially obtained with respect to the trend data, and the movement of the approximated curved surface is grasped as a change in a parameter depending on the time of the curved surface.

Const = c ₁ T ² + c ₂ T + c ₃ (2) A regression curve is obtained as a function of time based on a change in this parameter, and a future surface is predicted using the regression curve, thereby predicting a desired time. Get the value. By taking out the value of the required area from these, it is set as the predicted value.

Next, the function operation device 3 inputs the sunshine, the temperature, and the day of the week, and calculates the heat demand and the electricity demand of the next day by the function. This function is obtained by the learning device 4. The learning device 4 first acquires standard data as shown in FIG. 4 as an average demand pattern for each season and each day of the week. The pattern is represented as a non-linear mapping function by, for example, a spline function or a Bezier function. Next, demand data actually generated for each season and each day of the week is collected, and a table is created for the set using sunshine and temperature as parameters as shown in FIG. FIG. 4 shows, for example, a change in the standard data from the standard data, which is a characteristic diagram of the sunshine and the heat demand, according to the magnitude of the sunshine as a parameter and the level of the temperature.

In practice, a table is created for such standard data by expressing the spline parameters in a regression equation using sunshine and temperature as parameters. The regression equation obtained in this way is passed to the function operation device 3, and the sunshine, temperature prediction data,
In addition, heat and power demand forecast is performed for the date.

As described above, according to the present embodiment, the following operation is achieved. That is, in the composite utility that supplies heat, cold, and power to a specific area or building, in the present embodiment, the heat demand and the power demand, which have been predicted completely separately, are combined with the communication device 1 and the arithmetic device on one computer. Since the functions common to the heat demand and the power demand prediction are integrated into one, it is possible to significantly reduce the cost in terms of the system configuration.

Further, by adding a communication device 1 which has not been provided as a forecast function, accurate weather forecast data can be automatically used through a communication line, especially a telephone line, and a great demand forecast accuracy can be obtained. Improvements can be expected and operator demand can be reduced.

FIG. 5 shows another embodiment in which the communication device 1 directly obtains sunshine and temperature prediction data. By obtaining data supplied by the Meteorological Agency or the like through the communication device 1, it is possible to obtain sunshine / temperature prediction data. In this case, the most accurate prediction data can be obtained.
The communication device 1 obtains data by connecting a communication line for a predetermined time, for example, every three hours.

In the present embodiment, by providing the communication device 1 with a function of making a telephone call at regular time intervals, reading data, and then disconnecting the data, it is possible to greatly reduce communication costs.

[0029]

As described above, according to the present invention, weather-related data is taken in from outside by a communication device without having a power demand forecasting system and a heat demand forecasting system as in the prior art. By outputting at least one of the heat demand forecast value and the power demand forecast value based on the data, accurate weather forecast data can be automatically used through the communication line, and the demand forecast accuracy is greatly improved. Thus, it is possible to provide a thermoelectric demand forecasting system capable of optimizing the energy efficiency of an energy plant.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing one embodiment of a thermoelectric demand prediction system according to the present invention.

FIG. 2 is a diagram showing an example of AMeDAS data.

FIG. 3 is a diagram showing an example of an analysis result of AMeDAS data.

FIG. 4 is an explanatory diagram of a prediction function.

FIG. 5 is a configuration diagram showing another embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Communication apparatus 2 ... Image processing apparatus 3 ... Function calculation apparatus 4 ... Demand learning apparatus

Claims (11)

[Claims] 1. A thermoelectric demand forecasting system for an energy plant that supplies at least one of heat and electric power to at least one of heat and electric power to be supplied, a communication device that takes in weather-related data from outside, and the communication device. And at least one of a heat demand prediction and a power demand prediction based on the meteorological data captured by the computer, and outputting the result. 2. The communication device according to claim 1, wherein the communication device has means for acquiring weather-related data from an information source at regular intervals through a telephone line, and disconnecting the telephone line after acquiring the data. Thermoelectric demand forecasting system. 3. The thermoelectric demand forecasting system according to claim 2, wherein said communication device is connected to the Internet and has means for downloading data from a database connected via the Internet. 4. The thermoelectric demand forecasting system according to claim 2, wherein said communication device has means for acquiring at least AMeDAS data on sunshine and temperature. 5. The thermoelectric demand prediction system according to claim 2, wherein the communication device has a unit for acquiring a cloud image of at least a region including Japan. 6. The thermoelectric demand forecasting system according to claim 2, wherein said communication device has means for acquiring sunshine forecast data and temperature forecast data from an information source of weather-related data. 7. The demand learning means for generating a regression curve for associating the weather-related data obtained by the communication device with at least one of heat data and power demand data obtained by the communication device; 2. The thermoelectric demand forecasting system according to claim 1, further comprising an arithmetic unit for performing demand forecasting using a regression curve generated by the demand learning means. 8. The thermoelectric demand forecasting system according to claim 7, wherein the regression curve generated by the demand learning means is a nonlinear mapping function. 9. The thermoelectric demand forecasting system according to claim 8, wherein the nonlinear mapping function generated by the demand learning means is a spline function or a Bezier function. 10. The thermoelectric demand forecasting system according to claim 9, wherein said demand learning means expresses a demand characteristic depending on sunshine or temperature by a regression equation of a parameter of a generated spline function or Bezier function. . 11. The thermoelectric demand forecasting system according to claim 1, wherein the arithmetic unit has means for calculating the movement of sunshine and temperature from a distribution map of AMeDAS data in all over Japan or a specific area.