WO2019235272A1 - Energy-saving diagnostic program - Google Patents

Energy-saving diagnostic program Download PDF

Info

Publication number
WO2019235272A1
WO2019235272A1 PCT/JP2019/020788 JP2019020788W WO2019235272A1 WO 2019235272 A1 WO2019235272 A1 WO 2019235272A1 JP 2019020788 W JP2019020788 W JP 2019020788W WO 2019235272 A1 WO2019235272 A1 WO 2019235272A1
Authority
WO
WIPO (PCT)
Prior art keywords
value
electricity usage
data
electricity
time
Prior art date
Application number
PCT/JP2019/020788
Other languages
French (fr)
Japanese (ja)
Inventor
卓勇 山口
Original Assignee
備前グリーンエネルギー株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 備前グリーンエネルギー株式会社 filed Critical 備前グリーンエネルギー株式会社
Priority to JP2019566853A priority Critical patent/JP6756425B2/en
Publication of WO2019235272A1 publication Critical patent/WO2019235272A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B70/00—Technologies for an efficient end-user side electric power management and consumption
    • Y02B70/30—Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
    • Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P80/00—Climate change mitigation technologies for sector-wide applications
    • Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S20/00—Management or operation of end-user stationary applications or the last stages of power distribution; Controlling, monitoring or operating thereof
    • Y04S20/20—End-user application control systems
    • Y04S20/242—Home appliances
    • Y04S20/244—Home appliances the home appliances being or involving heating ventilating and air conditioning [HVAC] units

Definitions

  • the present invention relates to a program for automatically performing an energy saving diagnosis by using electricity usage data in units of buildings such as offices and factories and building address information.
  • Patent Literature 1 presents a program for diagnosing energy usage by use as a pre-stage for performing energy saving diagnosis, creating a model case using the building size and usage as input values, and using the model case By allocating the actual measured values of energy usage by the composition ratio of different energy usage, the energy usage by application is estimated.
  • Patent Document 2 creates a prediction model using past electricity usage data in units of buildings, weather data, and equipment operation status, and provides weather forecast data and equipment operation status during a forecast period. Presents a program that predicts electricity usage by application.
  • Patent Literature 3 inputs past weather data, performs multiple regression using electricity usage as an output, creates a prediction model, and predicts electricity usage by inputting weather forecast data into the prediction model. The method is presented. The type of data used for input is set in advance by the analyst.
  • Patent Document 1 since a model case is created using the building size and usage pattern as input values, if an appropriate model case cannot be set in a building where the building size and usage pattern are unknown, the energy usage by application The accuracy of the estimation is low. Moreover, in patent document 2, it is necessary to input the operating condition of the past installation in advance for the prediction of the amount of electricity used. Furthermore, if the operating status is not known, the analyst needs to input the operating status, but the same operating status assumption does not apply to various buildings. If it is different from the actual operating status, the accuracy of the prediction is low. In Patent Document 3, a multiple regression model is created from past weather data and electricity usage data, and weather forecast data is input to the multiple regression model to obtain prediction data.
  • Patent Document 3 provides a method for analyzing energy usage to an analyst, and automatically predicts usage using weather data and electricity usage data. Rather than do it, it requires the analyst to know in advance for energy usage analysis. Furthermore, for example, when using the technology of Patent Document 1, Patent Document 2 and Patent Document 3 for a customer's building in an electric power company or the like, the scale and usage pattern of the building for a large number of buildings such as tens of thousands It is not easy to acquire and input building-specific information such as working days and working hours. In view of the above problems, the present invention provides a program capable of energy saving diagnosis for a large number of buildings such as offices and factories by using easily available information without requiring specialized knowledge. For the purpose.
  • the present invention comprises the following technical means for the above purpose.
  • the first technical means is a computer, Using the building's electricity usage data measured at regular intervals from the outside and the weather data around the building, the use of the building at the measurement date and time, air conditioning, lighting, etc.
  • a program for calculating the amount of electricity used separately, It consists of a discrimination module, a model type estimation module, and an electricity usage calculation module for each application.
  • the discriminating module creates a discriminant model for discriminating between operation and non-operation at the measurement date and time by using the data for each measurement time zone obtained by extracting the electricity usage data and the weather data for each measurement time zone, In the measurement time zone, to determine the operation and non-operation at the target measurement date and time,
  • the model type estimation module uses the operation / non-operation determined by the determination module and the data for each measurement time zone, and uses the data for each measurement time zone for each measurement time zone, for each operation / non-operation.
  • the usage-specific electricity usage calculation module is based on the measurement time zone operation / non-operational electricity usage estimation model, the outside temperature of the weather data of the measurement time zone data, or the upper limit value of the outside temperature and humidity Within the range of the lower limit value, calculate the minimum value of the estimated electricity usage calculated from the above-mentioned operating / non-operating electricity usage estimation model by measurement time period as the baseline by operating time / non-operational period by measurement time period.
  • the difference between the electricity usage data and the baseline for each measurement time zone operation / non-operation is calculated as the amount of electricity used for air conditioning, Calculate the minimum value as the base electricity consumption in all the above-mentioned baselines by operation time / non-operation time, Calculate the value calculated as the difference between the baseline for each operation time / non-operation for each measurement time period and the amount of base electricity used as the amount of electricity used for lighting, etc. It is provided with a program for calculating the electricity usage estimation model by operation time / non-operation, the operation / non-operation flag, and the electricity usage by application such as air conditioning, lighting, etc.
  • the second technical means is the computer in the first technical means, Using the weather forecast value obtained from the outside, the weather data, and the electricity usage data, the electricity usage forecast value, the occurrence probability distribution of the electricity usage forecast value, air conditioning, lighting, etc. It is a program that calculates the electricity usage forecast value for each base application, Using the first technical means to calculate the operating / non-operating flag, the operating / non-operating baseline by measurement time zone, and the base electricity usage, The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operational baseline by measurement time zone, which is a set of measurement dates and times that are non-operational in the baseline by operation / non-operational time period, Set it up and modify the working / non-working flag, Enter the day of the week, holiday, time zone, etc.
  • an operation regression model with the modified operation / non-operation flag as output
  • In the operating regression model enter the day of the week, holiday and time zone to be predicted, calculate the operating / non-operating prediction flag, Regression analysis using the weather data as input and the electricity usage data as output for each modified operation / non-operation for each measurement time period, and creating an electricity usage prediction model for each operation time / operation period do it, Input the weather forecast value into the electricity usage amount prediction model for each operation time / non-operation time, calculate the electricity usage prediction value, and the probability distribution of the electricity usage prediction value, Using the electricity usage amount predicted value, the operation / non-operation prediction flag, the measurement time zone operation / non-operation baseline, and the base electricity usage amount, the electricity usage prediction value and the air conditioning Etc., and a program for calculating a predicted electricity usage amount for each application such as lighting.
  • the third technical means is the computer in the first or second technical means, A program that calculates the amount of energy saved by changing the temperature setting for air conditioning.
  • Set the change amount of air conditioning set temperature in advance Using the meteorological data, the electricity usage data, and the baseline by operation / non-operation by measurement time zone, As the temperature threshold at which the cooling and heating are switched, the temperature when the baseline by operation / non-operation is generated by the measurement time zone, When the temperature at the measurement date / time is higher than the temperature threshold value, the value obtained by subtracting the amount of change in the air conditioning set temperature from the temperature, and when the subtracted value is smaller than the temperature threshold value, the value is corrected to the temperature threshold value.
  • the value is the temperature for energy saving calculation.
  • the temperature for energy saving calculation Electricity for energy saving calculation calculated by inputting data obtained by replacing the temperature of the meteorological data with the temperature for energy saving calculation into the model for estimating electricity usage by operation time / non-operating time derived by the first technical means.
  • a program for calculating a difference between the estimated electricity usage value and the estimated electricity usage value for energy saving calculation as an energy saving amount by changing a set temperature of air conditioning is provided.
  • the fourth technical means is the computer according to any one of the first to third technical means, A program that calculates a reasonable target value for power demand, Predetermining the probability of occurrence of abnormal values in advance, Using the total value of the electricity usage data at 30-minute intervals and the average value of the weather data at 30-minute intervals as the technique of the first technical means, the electricity usage by operating time / non-operating time Calculate the estimation model, As a current power demand value, a value obtained by doubling the maximum value of the total value of 30 minutes intervals of the electricity usage data, As a demand value that can be achieved, a value obtained by doubling the maximum value of the electricity usage estimated value calculated from the electricity usage estimation model classified by operation / non-operation according to the measurement time period, When creating the electricity usage estimation model by operation time period / non-operation time, calculate the probability distribution of the estimated value, and calculate the estimated value from which the occurrence probability of the abnormal value becomes a threshold from the probability distribution Then, double the value as a demand value that is easy to achieve, When each demand value
  • a fifth technical means includes any one of the first to fourth technical means in a computer, A program that displays a breakdown of power demand, Using the total value of 30 minutes intervals of the electricity usage data and the average value of 30 minutes intervals of the meteorological data as the method of the first technical means, the electricity usage for each base application such as air conditioning, lighting, etc. Calculate the quantity, The total value of the electricity usage data at intervals of 30 minutes and the value of electric power demand value and the base usage such as lighting such as air conditioning etc. As another power demand value, Sort the power demand values in descending order, A program for displaying a breakdown by use of the power demand value using a stacked bar graph is provided.
  • Sixth technical means includes any one of the first to fifth technical means in a computer, A program that calculates additional electricity usage due to overtime work such as overtime,
  • the first technical means is used to calculate the operation / non-operation flag, the electricity usage estimation model for each operation time / non-operation time, and the baseline for each operation time / non-operation time by measurement time zone.
  • the measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are non-operational in the baseline by operation time / non-operation time, is set to non-operation. Correct the operating / non-operating flag, Enter the day of the week, holiday, time zone, etc.
  • Operation / Non-operation means that there are basically two situations in the building, the situation where the employee has joined the office and is working, the absence of employees, and the main equipment shut down In the case where the situation is non-operating and used for analysis, the operation / non-operation is handled as a flag.
  • the amount of electricity used for each application such as air conditioning, lighting, etc., “air conditioning etc.” mainly refers to the amount of electricity used for air conditioning.
  • Lighting refers mainly to the amount of electricity used by lighting, OA equipment, and the like.
  • Base refers to the amount of electricity used by a device such as a guide light for 24 hours.
  • the measurement time zone refers to a value obtained by extracting the time zone of the measurement date and time of a periodic cycle for the electricity usage data of the building.
  • the energy saving diagnosis program of the present invention from the electricity usage data of the building and the building address data, the estimation of the electricity usage by usage of the building, the prediction of the electricity usage, the prediction of the electricity usage by usage, the energy saving analysis, etc. It is possible to extract buildings that are easy to save energy from many offices and factories.
  • the energy-saving diagnosis program of the present invention does not require special technical knowledge related to energy-saving diagnosis, etc., and energy-saving diagnosis is achieved by using easily obtainable data such as building electricity usage data and building address data. There is an advantage that you can.
  • a building having a high energy saving effect can be efficiently extracted from a large number of buildings by carrying out an energy saving diagnosis for a customer building such as an electric power company.
  • the energy-saving diagnosis program of the present invention is based on the past electricity usage, past air conditioning, etc. lighting, etc., based on the usage of each base, predicted usage of electricity, and air conditioning etc., such as lighting, etc. It is possible to display a graph of the usage forecast value and the probability distribution of the electricity usage forecast value, making it easy to find out the past occurrence of electricity usage abnormalities, and to grasp the transition of future electricity usage, which will be useful for energy saving measures There is an advantage that you can.
  • the energy saving diagnosis program of the present invention can display a breakdown of the power demand value for each base application, such as air conditioning, lighting, etc. When a reasonable target value of the power demand value or the maximum power demand value occurs It is possible to confirm the breakdown of electricity usage by application and to use it for energy saving measures.
  • FIG. 1 is a block diagram illustrating an example of a configuration of an analysis unit of the processing apparatus according to the present embodiment.
  • FIG. 2 is a block diagram illustrating an example of the configuration of the energy saving diagnosis system according to the present embodiment.
  • FIG. 3 is a block diagram illustrating an example of the configuration of the processing apparatus according to the present embodiment.
  • FIG. 4 is a diagram illustrating an example of an analysis method for generating analysis data by combining electric usage data and weather data, creating a measurement time zone flag.
  • FIG. 5 is a diagram illustrating an example of a method for determining whether a building is operating / not operating for the determination module.
  • FIG. 6 is a flowchart showing an example of the operation of the model type estimation module.
  • FIG. 1 is a block diagram illustrating an example of a configuration of an analysis unit of the processing apparatus according to the present embodiment.
  • FIG. 2 is a block diagram illustrating an example of the configuration of the energy saving diagnosis system according to the present embodiment.
  • FIG. 3 is a
  • FIG. 7 is a diagram illustrating an example of a calculation technique of the usage-specific electricity usage calculation module.
  • FIG. 8 is a diagram illustrating an example of a calculation method of the usage-specific electricity usage calculation module.
  • FIG. 9 is a diagram illustrating an example of a graph display of a probability distribution of a past electricity usage amount, a past usage electricity usage amount, a predicted electricity usage amount value, a usage-specific electricity usage amount prediction value, and an electricity usage amount prediction value.
  • FIG. 10 is a diagram illustrating an example of a display of a demand curve graph of electricity usage by application.
  • FIG. 2 is a block diagram illustrating an example of the configuration of the energy saving diagnosis system 10 according to the present embodiment.
  • the energy saving diagnosis system 10 according to the present embodiment includes a client-side machine 3000, an external weather information database 4000, an external location information database 5000, an external communication network 6000, and a server 1000 as an analysis system.
  • the client side machine 3000 sends the electricity usage data 3002 and the building address data 3001 described later to the server 1000 as the analysis system through the external communication network 6000, and sends the analysis result of the analysis system to the client side machine 3000.
  • Meteorological observation facility location information 4001, meteorological observation data 4002, and meteorological forecast data 4003, which will be described later, are periodically sent from the external meteorological information database 4000 to the server 1000 as an analysis system through the external communication network 6000.
  • the building address data 3001 is sent to the external location information database 5000 through the external communication network 6000, and latitude, longitude, and altitude data necessary for analysis are obtained.
  • the building address data 3001 is information indicating a building address or zip code information.
  • the electricity usage data 3002 is electricity usage data measured at regular intervals (for example, at 30-minute intervals) from a smart meter or the like.
  • the weather observation facility location information data 4001 is latitude, longitude, and altitude data of the weather observation facility.
  • the meteorological observation data 4002 is data observed at a weather station such as temperature, humidity, sunshine duration, wind speed, wind direction, and precipitation.
  • the weather forecast data 4003 is data predicted at a weather station such as temperature, humidity, sunshine duration, wind speed, wind direction, and precipitation.
  • the position information database 5000 is latitude, longitude, altitude data, and the like.
  • the communication network 6000 is an Internet network, an intranet network, or the like.
  • the server 1000 includes, for example, the communication device 1 and the processing device 2.
  • the communication device 1 is an arbitrary device that is communicably connected to the external communication network 6000.
  • the processing device 2 executes various control processes in the server 1000.
  • the processing device 2 includes one or a plurality of servers including a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), an I / O interface (Input-Output Interface), and the like. Functions of each part are realized by causing a computer to execute a predetermined program.
  • FIG. 3 is a block diagram illustrating an example of the configuration of the processing apparatus 2 according to the present embodiment.
  • the processing apparatus 2 includes, for example, a communication processing unit 21, an input unit 22, a storage unit 23, an analysis unit 24, and an output unit 25.
  • the communication processing unit 21 controls the communication device 1 and acquires weather observation facility position information 4001, weather observation data 4002, weather forecast data 4003, building address data 3001, and electricity usage data 3002 via the external communication network 6000. To do.
  • the input unit 22 acquires the weather observation facility location information 4001, weather observation data 4002, weather forecast data 4003, building address data 3001, and electricity usage data 3002 from the communication processing unit 21, and the weather observation facility in the storage unit 23.
  • the database 231, the weather database 232, the building address database 233, and the electricity usage database 234 are stored in corresponding databases.
  • the analysis unit 24 acquires necessary data from the input unit 22 and the storage unit 23. Further, the analysis unit 24 is connected to the external position information database 5000 through the communication processing unit 21 through a Web API (Web Application Programming Interface) or the like, transmits building address data 3001, and building position information such as latitude, longitude, and altitude. To get. The configuration of the analysis unit 24 will be described in detail later.
  • FIG. 1 is a block diagram illustrating an example of the configuration of the analysis unit 24 of the processing apparatus 2 according to the present embodiment.
  • the analysis unit 24 includes, for example, a weather data acquisition package 241, a data preprocessing package 242, a usage-specific electricity usage estimation package 243, an electricity usage prediction package 244, and an energy saving analysis package 245.
  • a package is a set of program codes roughly classified by function, and a module is a set of program codes roughly classified by analysis content.
  • the weather data acquisition package 241 includes, for example, a position information acquisition module 2411 and a weather data extraction module 2412.
  • the location information acquisition module 2411 transmits the building address data 3001 of the building to be analyzed from the building address database 233 of the storage unit 23 through the communication processing unit 21 to the external location information database 5000 through the Web API or the like, and the latitude of the building, Acquire location information such as longitude and altitude.
  • a weather observation facility within the range of latitude ⁇ 1 degree and longitude ⁇ 1 degree of the building position information is extracted from the weather observation facility database 231 of the storage unit 23.
  • the distance between the two points of the building and the extracted weather observation facility is calculated, and the weather observation facility closest to the building is selected.
  • the meteorological data extraction module 2412 executes the analysis of the meteorological data of the meteorological observation facility located closest to the building to be analyzed from the meteorological database 232 of the storage unit 23 in the electricity usage database 234 of the storage unit 23. To extract the electricity usage data and the same period. Weather forecast data is also extracted.
  • the data preprocessing package 242 acquires electricity usage data, weather data, and weather forecast data.
  • the data preprocessing package 242 matches the electricity usage data, weather data, and weather forecast data at the same measurement interval. Basically, up-sampling or down-sampling is performed on weather data and weather forecast data in accordance with the measurement interval of electricity usage data. Upsampling uses linear interpolation, spline interpolation, or the like. Each data after the interpolation is analyzed using an abnormal value detection method, and if an abnormal value is found, it is deleted and treated as a missing data. If each data is missing for a short time, the data is interpolated by linear interpolation or the like. If it is a long-term deficit, leave it deficient. As shown in FIG.
  • the data preprocessing package 242 combines the electricity usage data 242c and the weather data 242d using the measurement date 242a as a key, and creates a measurement time zone flag 242b from the measurement date 242a.
  • the measurement time zone flag 242b is calculated as, for example, the value of time of measurement date ⁇ 100 + the value of minute of measurement date.
  • Electricity usage estimation package 243 by application It is a package that estimates electricity usage by application from analysis data.
  • the usage-specific electricity usage estimation package 243 includes, for example, a discrimination module 2431, a model type estimation module 2433, and a usage-specific electricity usage calculation module 2435.
  • the determination module 2431 is a module that determines whether the building is operating / not operating at the measurement date and time. About analysis data, operation / non-operation is discriminated from a set of a pair of electricity usage data and weather data. As a discrimination method, a method using a K-means algorithm, a method using a conditional mixture model, and the like can be considered. Here, a method using a support vector machine will be described. The method using the support vector machine requires training data for operation / non-operation. For training data for operation / non-operation, the range from the minimum value to the maximum value of the temperature of the analysis data is divided into 8 or 10 for example, and several points are counted from the maximum in the electricity usage data within the range.
  • FIG. 5 shows an example of a building operation / non-operation determination method for the determination module 2431.
  • the horizontal axis indicates the air temperature
  • the vertical axis indicates the electricity usage
  • the analysis data of the pair of electricity usage data and weather data is the white circle (2431a)
  • the operation training data is the black circle (2431b)
  • the non-operation training Data is indicated by a cross (2431c)
  • a dividing line when the range from the minimum value to the maximum value is divided into eight is indicated by a dotted line (2431d).
  • the model type estimation module 2433 creates an electricity usage amount estimation model for each operation time / non-operation time to estimate the electricity usage by using the data for each measurement time zone extracted from the analysis data for each measurement time zone. Yes, this will be described with reference to the flowchart of FIG.
  • the estimation of the amount of electricity used is started from the measurement time zone data (S201).
  • the processing from S202 to S209 is repeated for each measurement time period.
  • the data classified by measurement time zone is discriminated as active and non-operating using the discrimination module 2431 (S203 to S204).
  • Training data when a support vector machine is used for the discrimination module 2431 includes a method of dividing the temperature into a plurality of parts and operating the upper few points of the electricity usage data and disabling the lower several points, and the model type estimation module 2433. Before the operation, the period operation / non-operation flag extracted in the measurement time zone among the operation / non-operation flags created by the discrimination module 2341 using the weather data and the electricity usage data is used as training data. The method to be used is considered.
  • Extracts data by measurement time zone for each operation / non-operation inputs meteorological data, outputs electricity usage data, and performs polynomial regression analysis including regularization terms or kernel regression analysis for each measurement time zone
  • An electricity usage estimation model for each operation / non-operation is created (S205 to S208).
  • a probability distribution of predicted electric usage values is created for the operating / non-operating flag and the weather data, using the electric usage amount estimation model for each operating time / non-operating time. The probability distribution indicates an output generation probability with respect to a certain input.
  • the model type estimation module 2433 is executed, and the obtained results are combined (2432 to 2434).
  • the usage-specific electricity usage calculation module 2435 will be described with reference to FIGS.
  • the electricity usage calculation module 2435 for each application receives the electricity usage estimation model for each operation time / non-operation time and the operation / non-operation flag, and uses electricity such as air conditioning for each operation time / non-operation time. The amount, the electricity usage such as lighting, and the base electricity usage are calculated. Within the range of the meteorological data of the data by measurement time zone from the electricity usage estimation model by operation time / non-operation, the operation / The minimum value of estimated electricity usage by non-operation is set as the baseline by operation / non-operation by measurement time period.
  • FIG. 7 is a diagram illustrating an example of a method for determining the cooling range and the heating range of the electric usage amount.
  • the horizontal axis indicates the temperature
  • the vertical axis indicates the electricity usage
  • the electricity usage data during operation is a white circle (2438a)
  • the electricity usage data during non-operation is a cross (2438b)
  • the electricity during operation The boundary line between the cooling range and the heating range of the usage amount data is a dotted line 2438e
  • the boundary line of the cooling range and the heating range of the non-operating electricity usage data is a dotted line 2438f, which is the minimum value of the estimated electricity usage amount during the operation.
  • the base line is indicated by a dotted line 2438c, and the base line which is the minimum value of the estimated electricity consumption when not in operation is indicated by a dotted line 2438d.
  • the cooling range and the heating range are different for operation and non-operation.
  • the air temperature data and the electricity usage data are shown to have a significant positive correlation using the data for each operating time / non-operation in the measurement range of the cooling range, it is considered that there is cooling and the electricity usage
  • the difference between the data and the baseline for each operating time / non-operating time is the amount of electricity used for air conditioning. If it is not significant, the electricity usage amount such as air conditioning is set to 0, and the baseline of the corresponding measurement date is corrected to the electricity usage amount of the relevant measurement date.
  • the electricity usage data and The difference between the operating time and the non-operating baseline for each measurement time zone is the amount of electricity used for air conditioning. If it is not significant, the electricity usage amount such as air conditioning is set to 0, and the baseline of the corresponding measurement date is corrected to the electricity usage amount of the relevant measurement date. Of all the baselines by operating time / non-operating by measurement time zone before correction, the minimum baseline value is used as the base electricity consumption. At each measurement date and time, the electricity usage data is compared with the baseline. If the baseline is larger than the electricity usage data, the corresponding baseline is corrected to the electricity usage data and the electricity usage such as air conditioning.
  • the difference between the baseline and the base electricity usage is the electricity usage for lighting.
  • the measurement time zone flag, temperature, electricity usage, operation / non-operation, use / non-use of air conditioning, baseline, air conditioning, etc. It is an example which put together the electric consumption according to a use.
  • the baseline (non-operating: 1000) indicates the non-operating baseline in the measurement time zone 1000
  • the baseline (non-operating: 30) 'is the non-operating baseline in the measuring time zone 30 corrected.
  • the value indicates that the electricity usage [110] indicates that the electricity usage is 110 kWh.
  • the base usage-specific electricity usage such as operating / non-operating flag, air conditioning, etc.
  • the electricity usage prediction package 244 creates an electricity usage prediction model for each operation / non-operation by using the electricity usage data and the weather data, and the weather prediction value for the electricity usage prediction model for each operation / non-operation. Is used to calculate a predicted electricity usage amount, and the electricity usage amount for each base application such as air conditioning or lighting is predicted using the electricity usage prediction value. A probability distribution of predicted electricity usage is created, and the occurrence probability of the predicted electricity usage is calculated. The operating / non-operating flag is corrected using the calculation result of the usage-specific electricity usage estimation package 243.
  • the measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are not in operation in the baseline by operation time and non-operation time. , Correct the operation / non-operation flag.
  • An operation regression model is created by inputting the day of the week, holiday and time zone of the measurement date and time, and outputting the modified operation / non-operation flag.
  • Logistic regression analysis, support vector machine analysis, or the like can be used to create a regression model.
  • the operation / non-operation prediction flag is calculated by inputting the day of the week, the holiday, the time zone, etc. of the prediction date and time into the operation regression model.
  • a weather forecast value is input to the electricity usage amount prediction model for each operation time / non-operation time, and the electricity usage amount prediction value and the probability distribution of the electricity usage amount prediction value are calculated.
  • the electricity usage prediction model for each operation time / non-operation and the operation regression model are stored and used for estimation of electricity usage in real time.
  • the electricity usage prediction model and the operation regression model are updated periodically such as every month.
  • FIG. 9 shows past electricity usage data, past air conditioning, etc., lighting, etc., base usage of electricity, predicted usage, and air conditioning, lighting, etc., base usage, electricity usage prediction It is an example of the figure which displayed the value and the probability distribution of the electricity usage predicted value collectively as a graph.
  • a predicted electricity usage amount such as air conditioning
  • a predicted electricity usage amount such as lighting
  • a predicted base electricity usage amount such as lighting
  • FIG. 9 shows past electricity usage data, past air conditioning, etc., lighting, etc., base usage of electricity, predicted usage, and air conditioning, lighting, etc., base usage, electricity usage prediction It is an example of the figure which displayed the value and the probability distribution of the electricity usage predicted value collectively as a graph.
  • the horizontal axis indicates the date and time
  • the vertical axis indicates the electricity usage
  • the past electricity usage data 244i and the electricity usage predicted value 244j are shown in a line graph
  • the past base electricity usage 244c the lighting, etc.
  • the electric usage amount 244d and the electric usage amount 244e for air conditioning, etc. are shown in a stacked bar graph
  • the predicted value 244f for the base electric usage amount the predicted value 244g for the electric usage amount for lighting, etc.
  • the predicted value of the usage amount 244h and the predicted value of the electricity usage amount by use are shown in a stacked bar graph.
  • the probability that an abnormal value occurs is set in advance, and from the probability distribution of the electricity usage predicted value, a lower limit value at which the threshold value is 1 ⁇ 2 of the probability that the abnormal value occurs is set as a predicted lower limit value 244a.
  • An upper limit value at which the threshold value is 1 ⁇ 2 of the probability of occurrence of the abnormal value is shown as a predicted upper limit value 244b in a plane graph. For example, when the probability of occurrence of an abnormal value is set to 5%, the range between the prediction lower limit value 244a and the prediction upper limit value 244b is a range of prediction values that occur with a probability of 95% as a probability distribution.
  • the predicted electric usage amount, the predicted electric usage amount for each base such as air conditioning and lighting, and the probability distribution of the predicted electric usage amount are recorded in the analysis result database 235 of the storage unit 23.
  • the energy saving analysis package 245 performs an analysis that leads to energy saving using the calculation result of the usage-specific electricity usage estimation package 243.
  • the energy saving analysis package 245 includes, for example, an air conditioning set temperature module 2451, a demand analysis module 2452, and an after-hours analysis module 2453.
  • the air conditioning set temperature module 2451 calculates an energy saving amount when the air conditioning set temperature is changed.
  • the boundary line 2438e of the cooling range and heating range of the electricity usage data during operation shown in FIG. 7 is the temperature threshold during operation in the target measurement time zone, the cooling range and heating range of the electricity usage data during non-operation.
  • the boundary line 2438f is set as a temperature threshold value during non-operation in the target measurement time zone.
  • the temperature for energy saving calculation at the target measurement date / time is calculated using the temperature threshold value at which the operation / non-operation and the measurement time zone match at the measurement date / time. When the air temperature is higher than the air temperature threshold, a value obtained by subtracting the change amount of the air conditioning set temperature from the air temperature is used.
  • the temperature for energy saving calculation is set to the same value as the temperature threshold.
  • the air temperature is equal to or lower than the air temperature threshold, a value obtained by adding the change amount of the air conditioning set temperature to the air temperature is used.
  • the temperature for energy saving calculation is set to the same value as the temperature threshold.
  • Meteorological data obtained by replacing the temperature data with the temperature for energy saving calculation is input to the usage-specific electricity usage calculation module 2435 to estimate the energy usage for electricity saving calculation. For each measurement date and time, the difference between the estimated value of electricity usage calculated from past data and the electricity usage for energy saving calculation is calculated to obtain the energy saving amount by changing the setting.
  • the demand analysis module 2452 calculates a target value for a reasonable power demand value to facilitate reasonable power demand monitoring.
  • the occurrence probability of abnormal values is determined in advance.
  • the total value of the electricity usage data at the 30-minute interval and the average value of the weather data at the 30-minute interval are input to the usage-specific electricity usage estimation package 243.
  • a value obtained by doubling the maximum value of the total value of the electricity usage data every 30 minutes is set as the current power demand value.
  • Achievable demand value is obtained by doubling the maximum value of the estimated electricity usage calculated from the electricity usage estimation model by operation time / non-operation time derived by the electricity usage estimation package 243 by application. To do.
  • the probability distribution of the estimated value is calculated when the electricity usage amount estimation model for each operation time period is created. From the probability distribution, an estimated value having the occurrence probability of the abnormal value as a threshold value is calculated, and a value obtained by doubling the value is set as a demand value that can be easily achieved.
  • a value that is input to the air conditioning set temperature module 2451 and that is obtained by doubling the maximum value of the estimated electricity use amount for energy saving calculation is set as a demand value by changing the set temperature.
  • the achievable power demand value, the easily achievable power demand value, and the demand value by changing the set temperature are larger than the current power demand value, they are corrected to the current power demand value.
  • the horizontal axis represents time
  • the vertical axis represents power demand value
  • the amount of electricity used is combined with the amount of electricity used for each application, such as 24k for base electricity, 24l for electricity such as lighting, and 24m for electricity such as air conditioning. It is an example of the figure which rearranged data and displayed in the descending order of the amount of electricity used, and was displayed by the stacked bar graph.
  • the power demand value 24n indicates a current power demand value, an easily achieved power demand value, an achievable power demand value, and a power demand value by changing the set temperature.
  • the overtime analysis module 2453 calculates an additional amount of electricity used for overtime work such as overtime, and promotes energy saving by reducing overtime work. Calculate the operating / non-operating flag, the operating / non-operating electricity usage estimation model by measurement time zone, and the operating / non-operating baseline by measurement time zone using the usage-specific electricity usage estimation package 243 . The operating / non-operating flag is corrected using the calculation result of the usage-specific electricity usage estimation package 243.
  • the measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are not in operation in the baseline by operation time and non-operation time. , Correct the operation / non-operation flag.
  • An operation regression model is created by inputting the day of the week, holiday and time zone of the measurement date and time, and outputting the modified operation / non-operation flag. Logistic regression analysis, support vector machine analysis, or the like can be used to create a regression model. Input the day of the week, holiday, time zone, etc. of the measurement date and time into the operation regression model to calculate the regular operation / non-operation of the building.
  • Non-operating electricity usage at the measurement date and time of overtime work is estimated using the non-operating electricity usage estimation model by measurement time zone.
  • the difference between the amount of electricity used at the measurement date and time when the overtime work flag is set and the amount of non-operating electricity used at the measurement date and time of overtime work is calculated as an additional amount of electricity used by overtime work.
  • the result calculated by the energy saving analysis package 245 is recorded in the analysis result database 235 of the storage unit 23.
  • a large number of buildings such as offices and factories can be obtained by using only easily available information such as the amount of electricity used in the building and the building address information without special technical knowledge.
  • Energy-saving diagnosis is possible for
  • the electricity usage and address information of the building required in this embodiment is information that can be easily obtained by the power company etc. for customers, sending detailed energy-saving diagnosis regularly, or in real time It can also be used for services that check the diagnosis results via the network.
  • by conducting energy-saving diagnosis for buildings of many customers such as offices and factories it is possible to extract buildings with high energy-saving effects and promote highly efficient energy-saving measures.
  • an example of a preferable form is shown, and implementation is possible in addition to the contents described.
  • the representative example was shown about the formula, method, etc. which were introduced here, it is not limited to these, Other formulas can be substituted.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • Public Health (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Water Supply & Treatment (AREA)
  • Power Engineering (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

As a technology for promoting energy saving, there has been no method that enables energy saving diagnosis for a large number of buildings, such as offices and factories, using easily available information without specialized expertise. The present invention provides a program that, in a processing device for energy-saving diagnosis, undertakes such energy-saving diagnosis as estimation of basic electricity usage by intended use such as air conditioning and illumination, prediction of electricity usage, prediction of basic electricity usage by intended use such as air conditioning and illumination, and energy-saving analysis, from electricity usage data of relevant buildings and weather data acquired using building address data.

Description

省エネ診断プログラムEnergy saving diagnostic program
　本発明は、オフィス、工場等の建物単位の電気使用量データと、建物の住所情報と、を用いて、自動的に省エネ診断をするプログラムに関するものである。 The present invention relates to a program for automatically performing an energy saving diagnosis by using electricity usage data in units of buildings such as offices and factories and building address information.
　近年、省エネルギー（以下、省エネという。）を促進する機運が益々旺盛になってきており、その一環として、建物の省エネ診断（省エネ化のための診断）が行われている。そこで、簡易で精度の高い省エネ診断手法の開発と、効率良く省エネ対策支援を行うために、電力会社等で入手可能な情報を用いて、多数の建物の中から省エネ効果の高い建物を抽出する手法の開発が望まれている。
　建物の省エネ化を行う上で、建物のエネルギー使用状況を把握することは極めて重要であるが、近年、スマートメーターの普及が急激に進み、多くの建物で計測間隔が３０分間隔以下の電気使用量データを容易に入手できるようになっている。そのため、スマートメーター等で計測した電気使用量を用いて、プログラムにより省エネ診断を行ったり、電気使用量の予測を行う従来技術がある。
　例えば、特許文献１では、省エネ診断を行う前段として、用途別エネルギー使用量を診断するプログラムを提示しており、建物の規模及び使用形態を入力値として、モデルケースを作成し、モデルケースの用途別エネルギー使用量の構成比で、エネルギー使用量の実測値を按分することで、用途別エネルギー使用量を推計している。
　特許文献２は、建物単位の過去の電気使用量データと、気象データと、設備の稼働状況と、を用いて予測モデルを作成し、気象予報データと、予測する期間の設備の稼働状況と、から用途別電気使用量を予測するプログラムを提示している。
　特許文献３は、過去の気象データを入力して、電気使用量を出力とした重回帰を行い、予測モデルを作成し、予測モデルに気象予報データを入力することで、電気使用量を予測する方法を提示している。入力に用いるデータの種類は、事前に分析者が設定する。 In recent years, the momentum for promoting energy saving (hereinafter referred to as energy saving) has become more and more vigorous, and as part of this, energy saving diagnosis of buildings (diagnosis for energy saving) is performed. Therefore, in order to develop a simple and highly accurate energy-saving diagnosis method and efficiently support energy-saving measures, use the information available from electric power companies, etc., to extract buildings with high energy-saving effects from many buildings Development of a method is desired.
In order to save energy in buildings, it is extremely important to grasp the energy usage status of buildings. However, in recent years, the spread of smart meters has advanced rapidly, and in many buildings, electricity is used with a measurement interval of 30 minutes or less. Quantity data can be easily obtained. For this reason, there is a conventional technique for performing an energy saving diagnosis or predicting the amount of electricity used by a program using the amount of electricity measured by a smart meter or the like.
For example, Patent Literature 1 presents a program for diagnosing energy usage by use as a pre-stage for performing energy saving diagnosis, creating a model case using the building size and usage as input values, and using the model case By allocating the actual measured values of energy usage by the composition ratio of different energy usage, the energy usage by application is estimated.
Patent Document 2 creates a prediction model using past electricity usage data in units of buildings, weather data, and equipment operation status, and provides weather forecast data and equipment operation status during a forecast period. Presents a program that predicts electricity usage by application.
Patent Literature 3 inputs past weather data, performs multiple regression using electricity usage as an output, creates a prediction model, and predicts electricity usage by inputting weather forecast data into the prediction model. The method is presented. The type of data used for input is set in advance by the analyst.
特開２０１３−６５０８７号公報JP2013-65087A 特開２０１７−５０９７１号公報JP 2017-50971 A 特開２００４−１６４３８８号公報JP 2004-164388 A
　しかしながら、特許文献１では、建物の規模及び使用形態を入力値として、モデルケースを作成するため、建物の規模及び使用形態がわからない建物では、適切なモデルケースを設定できない場合、用途別エネルギー使用量の推計の精度は低くなる。
　また、特許文献２では、電気使用量の予測には、過去の設備の稼働状況を事前に入力する必要がある。さらに、稼働状況がわからなければ、分析者が稼働状況を仮定して入力する必要があるが、様々な建物に対して、同一稼働状況の仮定が適応するわけではなく、稼働状況の仮定が、実際の稼働状況と異なった場合は、予測の精度が低くなる。
　そして、特許文献３では、過去の気象データと電気使用量データから重回帰モデルを作成し、重回帰モデルに気象予報データを入力して、予測データを取得するものであるが、気象データの中で、例えば気温等、何を用いて、重回帰モデルを作成するかは、分析者が設定する必要がある。さらに、建物の稼働スケジュールに関わる情報は、別途入力する必要があり、加えて、建物の稼働スケジュールを用いて分析するかどうかも分析者が選択する必要がある。すなわち、特許文献３の技術は、分析者に対してエネルギー使用量を分析するための手法を提供しているものであり、気象データと電気使用量データを用いて自動的に使用量の予測を行うものではなく、分析者に、エネルギー使用量分析のための知識を、前もって要求するものである。
　さらに、例えば電力会社等で、顧客の建物に対して、特許文献１、特許文献２及び特許文献３の技術を利用する場合、数万件といった多数の建物を対象に、建物の規模や使用形態、稼働日、稼働時間等の建物固有の情報を取得し、入力することは容易でない。
　本発明は、前記の課題に鑑み、専門的な知識を必要とせず、容易に入手可能な情報を用いて、オフィス、工場等の多数の建物に対して、省エネ診断が可能なプログラムを提供することを目的とする。 However, in Patent Document 1, since a model case is created using the building size and usage pattern as input values, if an appropriate model case cannot be set in a building where the building size and usage pattern are unknown, the energy usage by application The accuracy of the estimation is low.
Moreover, in patent document 2, it is necessary to input the operating condition of the past installation in advance for the prediction of the amount of electricity used. Furthermore, if the operating status is not known, the analyst needs to input the operating status, but the same operating status assumption does not apply to various buildings. If it is different from the actual operating status, the accuracy of the prediction is low.
In Patent Document 3, a multiple regression model is created from past weather data and electricity usage data, and weather forecast data is input to the multiple regression model to obtain prediction data. Thus, it is necessary for the analyst to determine what is used to create the multiple regression model, such as temperature. Furthermore, it is necessary to separately input information related to the building operation schedule, and in addition, the analyst needs to select whether to analyze using the building operation schedule. That is, the technique of Patent Document 3 provides a method for analyzing energy usage to an analyst, and automatically predicts usage using weather data and electricity usage data. Rather than do it, it requires the analyst to know in advance for energy usage analysis.
Furthermore, for example, when using the technology of Patent Document 1, Patent Document 2 and Patent Document 3 for a customer's building in an electric power company or the like, the scale and usage pattern of the building for a large number of buildings such as tens of thousands It is not easy to acquire and input building-specific information such as working days and working hours.
In view of the above problems, the present invention provides a program capable of energy saving diagnosis for a large number of buildings such as offices and factories by using easily available information without requiring specialized knowledge. For the purpose.
　本発明は、前記目的のために、以下のような各技術手段から構成されている。
　第１の技術手段は、コンピュータに、
　外部から得られる定期的な周期で計測された建物の電気使用量データと前記建物周辺の気象データを用いて、計測日時における前記建物の稼働／非稼働と、空調等、照明等、ベースの用途別電気使用量と、を計算させるプログラムであって、
　判別モジュールと、モデル式推計モジュールと、用途別電気使用量計算モジュールと、から構成され、
　前記判別モジュールは、計測時間帯別に、前記電気使用量データと前記気象データを抽出した計測時間帯別データを用いて、計測日時における稼働と、非稼働と、を判別する判別モデルを作成し、前記計測時間帯における、前記対象計測日時における稼働と、非稼働と、の判別を実施して、
　前記モデル式推計モジュールは、前記判別モジュールで判別した稼働／非稼働と、前記計測時間帯別データと、を用いて、前記計測時間帯別に、稼働／非稼働別に、前記計測時間帯別データの気象データを入力、電気使用量データを出力とした回帰分析を行い、計測時間帯別稼働／非稼働別電気使用量推計モデルを作成して、
　前記用途別電気使用量計算モジュールは、前記計測時間帯別稼働／非稼働別電気使用量推計モデルから、前記計測時間帯別データの気象データの外気温又は、外気温と湿度との、上限値から下限値の範囲内において、前記計測時間帯別稼働／非稼働別電気使用量推計モデルから計算された電気使用量推計値の最小値を計測時間帯別稼働／非稼働別ベースラインとして計算して、
　前記計測時間帯において、稼働／非稼働別に、前記電気使用量データと、前記計測時間帯別稼働／非稼働別ベースラインと、の差を、空調等電気使用量として計算して、
　全ての前記計測時間帯別稼働／非稼働別ベースラインの中で、最小の値をベース電気使用量として計算して、
　前記計測時間帯別稼働／非稼働別ベースラインと、前記ベース電気使用量と、の差をそれぞれ計算した値を、照明等電気使用量として計算して、
　前記計測時間帯別稼働／非稼働別電気使用量推計モデルと、稼働／非稼働フラグと、空調等、照明等、ベースの用途別電気使用量と、を計算させるプログラムを備えていることを特徴とする。
　第２の技術手段は、第１の技術手段において、コンピューターに、
　外部から入手する気象予報値と、前記気象データと、前記電気使用量データと、を用いて、電気使用量予測値と、前記電気使用量予測値の発生確率分布と、空調等、照明等、ベースの用途別電気使用量予測値と、を計算させるプログラムであって、
　前記第１の技術手段を用いて前記稼働／非稼働フラグと、前記計測時間帯別稼働／非稼働別ベースラインと、前記ベース電気使用量と、を計算して、
　前記計測時間帯別稼働／非稼働別ベースラインの中で非稼働である計測日時の集合である前記計測時間帯別非稼働ベースラインの最大値より、前記電気使用量が小さい計測日時は、非稼働にして、前記稼働／非稼働フラグを修正して、
　前記計測日時の曜日、祝日、時間帯等を入力、前記修正した稼働／非稼働フラグを出力とした稼働回帰モデルを作成して、
　前記稼働回帰モデルに、予測する日時の曜日、祝日、時間帯等を入力して、稼働／非稼働予測フラグを計算して、
　計測時間帯別に、前記修正した稼働／非稼働別に、前記気象データを入力、前記電気使用量データを出力とした回帰分析を行い、計測時間帯別稼働／非稼働別電気使用量予測モデルを作成して、
　前記計測時間帯別稼働／非稼働別電気使用量予測モデルに、前記気象予報値を入力して、電気使用量予測値と、前記電気使用量予測値の確率分布と、を計算して、
前記電気使用量予測値と、前記稼働／非稼働予測フラグと、前記計測時間帯別稼働／非稼働別ベースラインと、前記ベース電気使用量と、を用いて、電気使用量予測値と、空調等、照明等、ベースの用途別電気使用量予測値と、を計算させるプログラムを備えていることを特徴とする。
　第３の技術手段は、第１または第２の技術手段において、コンピューターに、
　空調の設定温度の変更による省エネ量を計算させるプログラムであって、
　事前に空調設定温度の変化量を設定しておき、
　前記気象データと、前記電気使用量データと、前記計測時間帯別稼働／非稼働別ベースラインと、を用いて、
　前記計測時間帯別稼働／非稼働別ベースラインが発生した時の気温を冷房と暖房が切り替わる気温閾値として、
　前記気温閾値より計測日時の気温が高い場合は、前記気温から前記空調設定温度の変化量を引いた値で、引いた値が前記気温閾値より小さくなる場合は、値を前記気温閾値に補正した値を省エネ計算用気温として、
　前記気温閾値より前記計測日時の気温が低い場合は、前記気温から前記空調設定温度の変化量を加えた値で、加えた値が前記気温閾値より大きくなる場合は、値を前記気温閾値に補正した値を省エネ計算用気温として、
　第１の技術手段で導出した前記計測時間帯別稼働／非稼働別電気使用量推計モデルに、前記気象データの気温を前記省エネ計算用気温に置換したデータを入力して計算した省エネ計算用電気使用量推計値と、
　第１の技術手段で導出した前記計測時間帯別稼働／非稼働別電気使用量推計モデルに、前記気象データを入力して計算した電気使用量推計値を用いて、
前記電気使用量推計値と、前記省エネ計算用電気使用量推計値と、の差を、空調の設定温度の変更による省エネ量として計算させるプログラムを備えていることを特徴とする。
　第４の技術手段は、第１から第３のいずれか１の技術手段において、コンピューターに、
　合理的な電力デマンドの目標値を計算させるプログラムであって、
　事前に、異常値の発生確率を定めておき、
　前記電気使用量データの３０分間隔の合計値と、前記気象データの３０分間隔の平均値と、を第１の技術手段の手法に用いて、計測時間帯別稼働／非稼働別電気使用量推計モデルを計算し、
　前記電気使用量データの３０分間隔の合計値の最大値を２倍した値を、現状の電力デマンド値として、
　前記計測時間帯別稼働／非稼働別電気使用量推計モデルから計算される電気使用量推計値の最大値を２倍した値を、達成可能なデマンド値として、
　前記計測時間帯別稼働／非稼働別電気使用量推計モデルを作成する際に、推計値の確率分布を計算して、前記確率分布から、前記異常値の発生確率が閾値となる推計値を計算し、その値を２倍した値を、達成容易なデマンド値として、
　それぞれのデマンド値が前記現状の電力デマンド値より大きい場合は、前記現状の電力デマンド値に補正して計算させるプログラムを備えていることを特徴とする。
　第５の技術手段は、第１から第４のいずれか１の技術手段において、コンピューターに、
　電力デマンドの内訳を表示させるプログラムであって、
　前記電気使用量データの３０分間隔の合計値と、前記気象データの３０分間隔の平均値と、を第１の技術手段の手法に用いて、空調等、照明等、ベースの用途別電気使用量を計算し、
　前記電気使用量データの３０分間隔の合計値及び、前記空調等、照明等、ベースの用途別電気使用量をそれぞれ２倍した値を、電力デマンド値及び、空調等、照明等、ベースの用途別電力デマンド値として、
　前記電力デマンド値を降順にソートして、
　積み上げ棒グラフを用いて、前記電力デマンド値の用途別内訳を表示させるプログラムを備えていることを特徴とする。
　第６の技術手段は、第１から第５のいずれか１の技術手段において、コンピューターに、
　残業等の所定時間外労働による追加的な電気使用量を計算させるプログラムであって、
　第１の技術手段を用いて前記稼働／非稼働フラグと、前記計測時間帯別稼働／非稼働別電気使用量推計モデルと、前記計測時間帯別稼働／非稼働別ベースラインと、を計算して、
　前記計測時間帯別稼働／非稼働別ベースラインの中で非稼働である計測日時の集合である計測時間帯別非稼働ベースラインの最大値より、電気使用量が小さい計測日時は、非稼働にして、稼働／非稼働フラグを修正して、
　前記計測日時の曜日、祝日、時間帯等を入力、前記修正した稼働／非稼働フラグを出力とした稼働回帰モデルを作成して、
　前記稼働回帰モデルに、前記計測日時の曜日、祝日、時間帯等を入力して、建物の定期的な稼働／非稼働フラグを計算して、
　前記定期的な稼働／非稼働フラグが、非稼働である計測日時において、前記修正した稼働／非稼働フラグが稼働である場合に、所定時間外労働としてフラグを立てて、
　前記計測時間帯別非稼働電気使用量推計モデルを用いて、前記所定時間外労働の計測日時における非稼働電気使用量を推計して、
前記所定時間外労働フラグが立っている計測日時における電気使用量と、前記所定時間外労働の計測日時における非稼働電気使用量と、の差を、所定時間外労働による追加的な電気使用量として計算させるプログラムを備えていることを特徴とする。
　また、本明細書に用いられる用語は、次のように定義される。
　稼働／非稼働とは、建物には、基本的に２つの状況があると仮定し、従業員が出社し、労働している状況を稼働、従業員が不在で、主な設備が停止している状況を非稼働とし、分析に用いる場合は、稼働／非稼働をフラグとして扱う。
　空調等、照明等、ベースの用途別電気使用量とは、「空調等」は、主に空調電気使用量を指す。「照明等」は、主に照明やＯＡ機器等の電気使用量を指す。「ベース」は、誘導灯等の２４時間使用されている機器の電気使用量を指す。
　計測時間帯とは、建物の電気使用量データについて、定期的な周期の計測日時の時間帯を抽出した値を指す。 The present invention comprises the following technical means for the above purpose.
The first technical means is a computer,
Using the building's electricity usage data measured at regular intervals from the outside and the weather data around the building, the use of the building at the measurement date and time, air conditioning, lighting, etc. A program for calculating the amount of electricity used separately,
It consists of a discrimination module, a model type estimation module, and an electricity usage calculation module for each application.
The discriminating module creates a discriminant model for discriminating between operation and non-operation at the measurement date and time by using the data for each measurement time zone obtained by extracting the electricity usage data and the weather data for each measurement time zone, In the measurement time zone, to determine the operation and non-operation at the target measurement date and time,
The model type estimation module uses the operation / non-operation determined by the determination module and the data for each measurement time zone, and uses the data for each measurement time zone for each measurement time zone, for each operation / non-operation. Perform regression analysis with weather data as input and electricity usage data as output, and create electricity usage estimation model by operation time / non-operation time,
The usage-specific electricity usage calculation module is based on the measurement time zone operation / non-operational electricity usage estimation model, the outside temperature of the weather data of the measurement time zone data, or the upper limit value of the outside temperature and humidity Within the range of the lower limit value, calculate the minimum value of the estimated electricity usage calculated from the above-mentioned operating / non-operating electricity usage estimation model by measurement time period as the baseline by operating time / non-operational period by measurement time period. And
In the measurement time zone, for each operation / non-operation, the difference between the electricity usage data and the baseline for each measurement time zone operation / non-operation is calculated as the amount of electricity used for air conditioning,
Calculate the minimum value as the base electricity consumption in all the above-mentioned baselines by operation time / non-operation time,
Calculate the value calculated as the difference between the baseline for each operation time / non-operation for each measurement time period and the amount of base electricity used as the amount of electricity used for lighting, etc.
It is provided with a program for calculating the electricity usage estimation model by operation time / non-operation, the operation / non-operation flag, and the electricity usage by application such as air conditioning, lighting, etc. And
The second technical means is the computer in the first technical means,
Using the weather forecast value obtained from the outside, the weather data, and the electricity usage data, the electricity usage forecast value, the occurrence probability distribution of the electricity usage forecast value, air conditioning, lighting, etc. It is a program that calculates the electricity usage forecast value for each base application,
Using the first technical means to calculate the operating / non-operating flag, the operating / non-operating baseline by measurement time zone, and the base electricity usage,
The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operational baseline by measurement time zone, which is a set of measurement dates and times that are non-operational in the baseline by operation / non-operational time period, Set it up and modify the working / non-working flag,
Enter the day of the week, holiday, time zone, etc. of the measurement date and time, create an operation regression model with the modified operation / non-operation flag as output,
In the operating regression model, enter the day of the week, holiday and time zone to be predicted, calculate the operating / non-operating prediction flag,
Regression analysis using the weather data as input and the electricity usage data as output for each modified operation / non-operation for each measurement time period, and creating an electricity usage prediction model for each operation time / operation period do it,
Input the weather forecast value into the electricity usage amount prediction model for each operation time / non-operation time, calculate the electricity usage prediction value, and the probability distribution of the electricity usage prediction value,
Using the electricity usage amount predicted value, the operation / non-operation prediction flag, the measurement time zone operation / non-operation baseline, and the base electricity usage amount, the electricity usage prediction value and the air conditioning Etc., and a program for calculating a predicted electricity usage amount for each application such as lighting.
The third technical means is the computer in the first or second technical means,
A program that calculates the amount of energy saved by changing the temperature setting for air conditioning.
Set the change amount of air conditioning set temperature in advance,
Using the meteorological data, the electricity usage data, and the baseline by operation / non-operation by measurement time zone,
As the temperature threshold at which the cooling and heating are switched, the temperature when the baseline by operation / non-operation is generated by the measurement time zone,
When the temperature at the measurement date / time is higher than the temperature threshold value, the value obtained by subtracting the amount of change in the air conditioning set temperature from the temperature, and when the subtracted value is smaller than the temperature threshold value, the value is corrected to the temperature threshold value. The value is the temperature for energy saving calculation.
When the temperature at the measurement date and time is lower than the temperature threshold, a value obtained by adding the amount of change in the air conditioning set temperature to the temperature, and when the added value is greater than the temperature threshold, the value is corrected to the temperature threshold. As the temperature for energy saving calculation,
Electricity for energy saving calculation calculated by inputting data obtained by replacing the temperature of the meteorological data with the temperature for energy saving calculation into the model for estimating electricity usage by operation time / non-operating time derived by the first technical means. Usage estimates, and
Using the electricity usage estimation value calculated by inputting the weather data into the electricity usage estimation model by operation / non-operation by the measurement time period derived by the first technical means,
A program for calculating a difference between the estimated electricity usage value and the estimated electricity usage value for energy saving calculation as an energy saving amount by changing a set temperature of air conditioning is provided.
The fourth technical means is the computer according to any one of the first to third technical means,
A program that calculates a reasonable target value for power demand,
Predetermining the probability of occurrence of abnormal values in advance,
Using the total value of the electricity usage data at 30-minute intervals and the average value of the weather data at 30-minute intervals as the technique of the first technical means, the electricity usage by operating time / non-operating time Calculate the estimation model,
As a current power demand value, a value obtained by doubling the maximum value of the total value of 30 minutes intervals of the electricity usage data,
As a demand value that can be achieved, a value obtained by doubling the maximum value of the electricity usage estimated value calculated from the electricity usage estimation model classified by operation / non-operation according to the measurement time period,
When creating the electricity usage estimation model by operation time period / non-operation time, calculate the probability distribution of the estimated value, and calculate the estimated value from which the occurrence probability of the abnormal value becomes a threshold from the probability distribution Then, double the value as a demand value that is easy to achieve,
When each demand value is larger than the current power demand value, a program for correcting the current power demand value for calculation is provided.
A fifth technical means includes any one of the first to fourth technical means in a computer,
A program that displays a breakdown of power demand,
Using the total value of 30 minutes intervals of the electricity usage data and the average value of 30 minutes intervals of the meteorological data as the method of the first technical means, the electricity usage for each base application such as air conditioning, lighting, etc. Calculate the quantity,
The total value of the electricity usage data at intervals of 30 minutes and the value of electric power demand value and the base usage such as lighting such as air conditioning etc. As another power demand value,
Sort the power demand values in descending order,
A program for displaying a breakdown by use of the power demand value using a stacked bar graph is provided.
Sixth technical means includes any one of the first to fifth technical means in a computer,
A program that calculates additional electricity usage due to overtime work such as overtime,
The first technical means is used to calculate the operation / non-operation flag, the electricity usage estimation model for each operation time / non-operation time, and the baseline for each operation time / non-operation time by measurement time zone. And
The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are non-operational in the baseline by operation time / non-operation time, is set to non-operation. Correct the operating / non-operating flag,
Enter the day of the week, holiday, time zone, etc. of the measurement date and time, create an operation regression model with the modified operation / non-operation flag as output,
In the operation regression model, enter the day of the week, holiday, time zone, etc. of the measurement date and time, calculate the regular operation / non-operation flag of the building,
In the measurement date and time when the periodic operation / non-operation flag is non-operation, when the corrected operation / non-operation flag is operation, a flag is set as overtime for a predetermined time,
Using the non-operating electricity usage estimation model by measurement time zone, estimating the non-operating electricity usage at the measurement date of the overtime work,
The difference between the amount of electricity used at the measurement date and time when the overtime work flag is set and the amount of non-operating electricity used at the measurement date and time of the overtime work is defined as an additional amount of electricity used by overtime work. It is characterized by having a program for calculation.
Further, terms used in the present specification are defined as follows.
Operation / Non-operation means that there are basically two situations in the building, the situation where the employee has joined the office and is working, the absence of employees, and the main equipment shut down In the case where the situation is non-operating and used for analysis, the operation / non-operation is handled as a flag.
The amount of electricity used for each application, such as air conditioning, lighting, etc., “air conditioning etc.” mainly refers to the amount of electricity used for air conditioning. “Lighting” refers mainly to the amount of electricity used by lighting, OA equipment, and the like. “Base” refers to the amount of electricity used by a device such as a guide light for 24 hours.
The measurement time zone refers to a value obtained by extracting the time zone of the measurement date and time of a periodic cycle for the electricity usage data of the building.
　本発明の省エネ診断プログラムによれば、建物の電気使用量データと建物の住所データから、建物の用途別電気使用量の推計、電気使用量の予測、用途別電気使用量の予測、省エネ分析等を行い、多数のオフィス、工場等の建物の中から省エネが行いやすい建物の抽出が可能である。
　本発明の省エネ診断プログラムは、省エネ診断等に関する特別な専門的知識は不要であり、建物の電気使用量データと建物の住所データという、容易に取得可能なデータを利用して、省エネ診断が達成できるという利点がある。
　また、本発明の省エネ診断プログラムによれば、電力会社等の顧客の建物を対象に、省エネ診断を実施することで、多数の建物の中から、省エネ効果の高い建物の抽出を効率的に行うことが可能である。
　本発明の省エネ診断プログラムは、過去の電気使用量と、過去の空調等、照明等、ベースの用途別電気使用量と、電気使用量予測値と、空調等、照明等、ベースの用途別電気使用量予測値と、電気使用量予測値の確率分布についてのグラフ表示を可能とし、過去の電気使用量の異常発生の発見や、将来の電気使用量の推移が把握しやすく、省エネ対策に役立てることができるという利点がある。
　本発明の省エネ診断プログラムは、電力デマンド値について、空調等、照明等、ベースの用途別に内訳を表示させることができ、電力デマンド値の合理的な目標値や、最大電力デマンド値が発生した際の用途別電気使用量の内訳を確認でき、省エネ対策に役立てることができるという利点がある。 According to the energy saving diagnosis program of the present invention, from the electricity usage data of the building and the building address data, the estimation of the electricity usage by usage of the building, the prediction of the electricity usage, the prediction of the electricity usage by usage, the energy saving analysis, etc. It is possible to extract buildings that are easy to save energy from many offices and factories.
The energy-saving diagnosis program of the present invention does not require special technical knowledge related to energy-saving diagnosis, etc., and energy-saving diagnosis is achieved by using easily obtainable data such as building electricity usage data and building address data. There is an advantage that you can.
In addition, according to the energy saving diagnosis program of the present invention, a building having a high energy saving effect can be efficiently extracted from a large number of buildings by carrying out an energy saving diagnosis for a customer building such as an electric power company. It is possible.
The energy-saving diagnosis program of the present invention is based on the past electricity usage, past air conditioning, etc. lighting, etc., based on the usage of each base, predicted usage of electricity, and air conditioning etc., such as lighting, etc. It is possible to display a graph of the usage forecast value and the probability distribution of the electricity usage forecast value, making it easy to find out the past occurrence of electricity usage abnormalities, and to grasp the transition of future electricity usage, which will be useful for energy saving measures There is an advantage that you can.
The energy saving diagnosis program of the present invention can display a breakdown of the power demand value for each base application, such as air conditioning, lighting, etc. When a reasonable target value of the power demand value or the maximum power demand value occurs It is possible to confirm the breakdown of electricity usage by application and to use it for energy saving measures.
　図１は、本実施形態に係る処理装置の分析部の構成の一例を示すブロック図である。
　図２は、本実施形態に係る省エネ診断システムの構成の一例を示すブロック図である。
　図３は、本実施形態に係る処理装置の構成の一例を示すブロック図である。
　図４は、電気使用量データと気象データとを結合し、計測時間帯フラグを作成して分析データを作成する分析手法の一例を示す図である。
　図５は、判別モジュールについて、建物の稼働／非稼働の判別手法の一例を示す図である。
　図６は、モデル式推計モジュールの動作の一例を示すフローチャートである。
　図７は、用途別電気使用量計算モジュールの計算手法の一例を示す図である。
　図８は、用途別電気使用量計算モジュールの計算手法の一例を示す図である。
　図９は、過去の電気使用量、過去の用途別電気使用量、電気使用量予測値、用途別電気使用量予測値、電気使用量予測値の確率分布のグラフ表示の一例を示す図である。
　図１０は、用途別電気使用量のデマンドカーブグラフの表示の一例を示す図である。 FIG. 1 is a block diagram illustrating an example of a configuration of an analysis unit of the processing apparatus according to the present embodiment.
FIG. 2 is a block diagram illustrating an example of the configuration of the energy saving diagnosis system according to the present embodiment.
FIG. 3 is a block diagram illustrating an example of the configuration of the processing apparatus according to the present embodiment.
FIG. 4 is a diagram illustrating an example of an analysis method for generating analysis data by combining electric usage data and weather data, creating a measurement time zone flag.
FIG. 5 is a diagram illustrating an example of a method for determining whether a building is operating / not operating for the determination module.
FIG. 6 is a flowchart showing an example of the operation of the model type estimation module.
FIG. 7 is a diagram illustrating an example of a calculation technique of the usage-specific electricity usage calculation module.
FIG. 8 is a diagram illustrating an example of a calculation method of the usage-specific electricity usage calculation module.
FIG. 9 is a diagram illustrating an example of a graph display of a probability distribution of a past electricity usage amount, a past usage electricity usage amount, a predicted electricity usage amount value, a usage-specific electricity usage amount prediction value, and an electricity usage amount prediction value. .
FIG. 10 is a diagram illustrating an example of a display of a demand curve graph of electricity usage by application.
　以下、本発明における実施の具体的な形態について図面に基づき説明する。
　本実施形態に使用するデータの計測期間は、６か月以上のデータであれば分析可能である。
　図２は、本実施形態に係る省エネ診断システム１０の構成の一例について説明するブロック図である。
　本実施形態に係る省エネ診断システム１０は、クライアント側のマシン３０００と、外部気象情報データベース４０００と、外部位置情報データベース５０００と、外部通信ネットワーク６０００と、分析システムとしてのサーバー１０００を有する。
　クライアント側のマシン３０００から、外部通信ネットワーク６０００を通して、後記の電気使用量データ３００２及び建物住所データ３００１を、分析システムとしてのサーバー１０００に送り、分析システムで分析した結果をクライアント側のマシン３０００に送る。外部気象情報データベース４０００から、外部通信ネットワーク６０００を通して、後記の気象観測施設位置情報４００１、気象観測データ４００２、気象予報データ４００３を、定期的に分析システムとしてのサーバー１０００に送る。分析システムを導入したサーバー１０００から、外部通信ネットワーク６０００を通して、建物住所データ３００１を外部位置情報データベース５０００に送り、分析に必要な緯度、経度、高度データを入手する。
　建物住所データ３００１は、建物の住所又は郵便番号情報を示す情報である。
　電気使用量データ３００２は、スマートメーター等から定期的周期（例えば、３０分間隔）で計測された電気使用量データである。
　気象観測施設位置情報データ４００１は、気象観測施設の緯度、経度、高度データである。
　気象観測データ４００２は、気温、湿度、日照時間、風速、風向、降水量等の気象観測所で観測されたデータである。
　気象予報データ４００３は、気温、湿度、日照時間、風速、風向、降水量等の気象観測所で予測されたデータである。
　位置情報データベース５０００は、緯度、経度、高度データ等である。
　通信ネットワーク６０００は、インターネット網やイントラネット網等である。
　サーバー１０００は、例えば、通信機器１と、処理装置２を含む。通信機器１は、外部通信ネットワーク６０００と通信可能に接続する任意のデバイスである。
　処理装置２は、サーバー１０００における各種制御処理を実行する。処理装置２は、ＣＰＵ（Ｃｅｎｔｒａｌ　Ｐｒｏｃｅｓｓｉｎｇ　Ｕｎｉｔ）、ＲＡＭ（Ｒａｎｄｏｍ　Ａｃｃｅｓｓ　Ｍｅｍｏｒｙ）、ＲＯＭ（Ｒｅａｄ　Ｏｎｌｙ　Ｍｅｍｏｒｙ）、Ｉ／Ｏ　ｉｎｔｅｒｆａｃｅ（Ｉｎｐｕｔ−Ｏｕｔｐｕｔ　Ｉｎｔｅｒｆａｃｅ）等を具備する１つ又は複数のサーバーから構成され、コンピューターに所定のプログラムを実行させることで各部の機能を実現する。
　図３は、本実施形態に係る処理装置２の構成の一例を示すブロック図である。
　処理装置２は、例えば、通信処理部２１と、入力部２２と、記憶部２３と、分析部２４と、出力部２５を含む。
　通信処理部２１は、通信機器１を制御し、外部通信ネットワーク６０００を介して、気象観測施設位置情報４００１、気象観測データ４００２、気象予報データ４００３、建物住所データ３００１、電気使用量データ３００２を取得する。
　入力部２２は、通信処理部２１から、気象観測施設位置情報４００１、気象観測データ４００２、気象予報データ４００３、建物住所データ３００１、電気使用量データ３００２を取得し、記憶部２３にある気象観測施設データベース２３１、気象データベース２３２、建物住所データベース２３３、電気使用量データベース２３４の対応する各データベースに格納する。
　分析部２４は、入力部２２及び記憶部２３から必要データを取得する。さらに、分析部２４は、通信処理部２１を通して、外部位置情報データベース５０００にＷｅｂ　ＡＰＩ（Ｗｅｂ　Ａｐｐｌｉｃａｔｉｏｎ　Ｐｒｏｇｒａｍｍｉｎｇ　Ｉｎｔｅｒｆａｃｅ）等で接続し、建物住所データ３００１を送信し、緯度、経度、高度等の建物位置情報を取得する。分析部２４の構成については、後で詳しく説明する。分析部２４で分析した結果は、記憶部２３にある分析結果データベース２３５に格納される。
　出力部２５は、分析部２４から必要なデータを取得し、通信処理部２１を通して、クライアント側のマシン３０００で分析結果を出力する。
　建物住所データ３００１や電気使用量データ３００２は、外部からの入力が必要なく、既にデータベースに保管されている場合も想定される。
　入力部２２、分析部２４、出力部２５は、物理的に別々のサーバーである場合や、同一サーバーである場合も想定される。
　分析結果の出力は、出力部２５を通して、クライアント側のマシン３０００で表示される場合や、分析結果データベース２３５上で分析に利用される場合も想定される。
　図１は、本実施形態に係る処理装置２の分析部２４の構成の一例を示すブロック図である。
　分析部２４は、例えば、気象データ取得パッケージ２４１と、データ前処理パッケージ２４２と、用途別電気使用量推計パッケージ２４３と、電気使用量予測パッケージ２４４と、省エネ分析パッケージ２４５を含む。
　パッケージとは、大まかに機能毎に分類したプログラムコードの集合であり、モジュールとは、大まかに分析内容毎に分類したプログラムコードの集合である。
　気象データ取得パッケージ２４１は、例えば、位置情報取得モジュール２４１１と、気象データ抽出モジュール２４１２を含む。
　位置情報取得モジュール２４１１は、記憶部２３の建物住所データベース２３３から分析を行う建物の建物住所データ３００１を、通信処理部２１を通して、外部位置情報データベース５０００にＷｅｂ　ＡＰＩ等を通じて送信し、建物の緯度、経度、高度等の位置情報を取得する。取得した建物の位置情報を用いて、記憶部２３の気象観測施設データベース２３１から、例えば建物位置情報の緯度±１度、経度±１度の範囲内にある気象観測施設を抽出する。建物と、前記の抽出した気象観測施設と、の２点間の距離を計算して、建物に最も近い位置にある気象観測施設を選択する。
　気象データ抽出モジュール２４１２は、記憶部２３の気象データベース２３２から、分析を実行する建物に最も近い位置にある気象観測施設の気象データを、記憶部２３の電気使用量データベース２３４のうち、分析を実行する電気使用量データと同期間分抽出する。また、気象予報データも抽出する。
　データ前処理パッケージ２４２は、電気使用量データと、気象データと、気象予報データと、を取得する。
　データ前処理パッケージ２４２は、電気使用量データと、気象データと、気象予報データと、を同じ計測間隔に合わせる。基本的には、電気使用量データの計測間隔に合わせて、気象データと、気象予報データと、に対してアップサンプリング又はダウンサンプリングを行う。アップサンプリングは、線形補間やスプライン補間等を用いる。補間後の各データは、異常値検知の手法を用いて分析し、異常値が発見された場合は、削除して欠損扱いにする。各データの欠損は、短期間の欠損であれば、直線補間等によりデータを補間する。長期間の欠損であれば、欠損のままとする。
　図４に示すように、データ前処理パッケージ２４２は、計測日時２４２ａをキーとして、電気使用量データ２４２ｃと、気象データ２４２ｄと、を結合し、計測日時２４２ａから、計測時間帯フラグ２４２ｂを作成して、分析データを作成する。計測時間帯フラグ２４２ｂは、例えば、計測日時の時間の値×１００＋計測日時の分の値のように計算する。
　用途別電気使用量推計パッケージ２４３は、
　分析データから、用途別電気使用量を推計するパッケージである。用途別電気使用量推計パッケージ２４３は、例えば、判別モジュール２４３１と、モデル式推計モジュール２４３３と、用途別電気使用量計算モジュール２４３５を含む。
　判別モジュール２４３１は、計測日時における建物の稼働／非稼働を判別するモジュールである。
　分析データについて、電気使用量データと気象データのペアの集合から、稼働／非稼働を判別する。判別手法としては、Ｋ−ｍｅａｎｓアルゴリズムを用いた方法や条件付き混合モデルを用いた方法等が考えられるが、ここではサポートベクトルマシンを用いた方法を説明する。
　サポートベクトルマシンを用いた方法は、稼働／非稼働の訓練データを必要とする。稼働／非稼働の訓練データは、分析データの気温の最小値から最大値の範囲を、例えば、８分割や１０分割等して、当該範囲内の電気使用量データで最大から数えて数点を訓練データの稼働、最小から数えて数点を訓練データの非稼働として抽出する。訓練データからサポートベクトルマシンにより判別モデルを作成して、残りの分析データを、判別モデルを用いて、稼働／非稼働に判別する。
　図５に、判別モジュール２４３１について、建物の稼働／非稼働の判別手法の一例を示す。
　図５で、横軸は気温、縦軸は電気使用量を示し、電気使用量データと気象データのペアの分析データは白丸（２４３１ａ）、稼働の訓練データは黒丸（２４３１ｂ）、非稼働の訓練データはバツ印（２４３１ｃ）、気温の最小値から最大値の範囲を８分割した場合の分割線は点線（２４３１ｄ）で示す。
　モデル式推計モジュール２４３３は、分析データを計測時間帯別に抽出した計測時間帯別データを用いて、電気使用量を推計する計測時間帯別稼働／非稼働別電気使用量推計モデルを作成するものであり、図６のフローチャートで説明する。
　計測時間帯別データから電気使用量の推計を開始する（Ｓ２０１）。
　計測時間帯毎にＳ２０２からＳ２０９の処理を繰り返す。
　当該計測時間帯別データを、判別モジュール２４３１を用いて、稼働と、非稼働と、に判別する（Ｓ２０３からＳ２０４）。
　判別モジュール２４３１にサポートベクトルマシンを用いた場合の訓練データは、気温を複数分割して、電気使用量データの上位数点を稼働、下位数点を非稼働とする方法と、モデル式推計モジュール２４３３を実施する前に、気象データ、電気使用量データを用いて、判別モジュール２３４１で作成した稼働／非稼働フラグの中で、当該計測時間帯で抽出された期間稼働／非稼働フラグを訓練データに用いる方法とが考えられる。
　稼働／非稼働別に計測時間帯別データを抽出して、気象データを入力、電気使用量データを出力とした、正則化項を含めた多項式回帰分析又は、カーネル回帰分析を行い、計測時間帯別稼働／非稼働別電気使用量推計モデルを作成する（Ｓ２０５からＳ２０８）。
　前記の計測時間帯別稼働／非稼働別電気使用量推計モデルを用いて、稼働／非稼働フラグと、気象データと、に対する、電気使用量予測値の確率分布を作成する。確率分布は、ある入力に対する、出力の発生確率を示す。
　計測時間帯別に、モデル式推計モジュール２４３３を実行し、得られた結果を結合する（２４３２から２４３４）。
　用途別電気使用量計算モジュール２４３５について、図７及び図８を用いて説明する。
　用途別電気使用量計算モジュール２４３５は、計測時間帯別稼働／非稼働別電気使用量推計モデルと、稼働／非稼働フラグと、を受け取り、計測時間帯別稼働／非稼働別に、空調等電気使用量と、照明等電気使用量と、ベース電気使用量と、を計算する。
　計測時間帯別稼働／非稼働別電気使用量推計モデルから、計測時間帯別データの気象データの範囲内において、計測時間帯別稼働／非稼働別電気使用量推計モデルから計算された、稼働／非稼働別の電気使用量推計値の最小値を、計測時間帯別稼働／非稼働別ベースラインとする。前記最小値を導出した気温を、気温閾値として、前記気温閾値より大きい気温の範囲を冷房範囲、気温閾値以下を暖房範囲とする。
　図７は、電気使用量の冷房範囲及び暖房範囲の判別手法の一例を示す図である。
　図７では、横軸は気温、縦軸は電気使用量を示し、稼働時の電気使用量データは白丸（２４３８ａ）、非稼働時の電気使用量データはバツ印（２４３８ｂ）、稼働時の電気使用量データの冷房範囲及び暖房範囲の境界線は点線２４３８ｅ、非稼働時の電気使用量データの冷房範囲及び暖房範囲の境界線は点線２４３８ｆ、稼働時の電気使用量推計値の最小値であるベースラインは点線２４３８ｃ、非稼働時の電気使用量推計値の最小値であるベースラインは点線２４３８ｄで示す。
　稼働と、非稼働と、で別々の冷房範囲、暖房範囲になる場合がある。冷房範囲の計測時間帯別稼働／非稼働別データを用いて、気温データと、電気使用量データと、に正の相関が有意にあると示された場合に、冷房はありとして、電気使用量データと、計測時間帯別稼働／非稼働別ベースラインと、の差を空調等電気使用量とする。有意でない場合は、空調等電気使用量を０として、更に該当する計測日時のベースラインを、該当する計測日時の電気使用量に補正する。同様に、暖房範囲の計測時間帯別稼働／非稼働別データを用いて、気温データと、電気使用量データと、に負の相関が有意にあると示された場合に、電気使用量データと、計測時間帯別稼働／非稼働別ベースラインと、の差を空調等電気使用量とする。有意でない場合は、空調等電気使用量を０として、更に該当する計測日時のベースラインを、該当する計測日時の電気使用量に補正する。
　全ての補正前の計測時間帯別稼働／非稼働別ベースラインの中で、最小のベースラインの値をベース電気使用量とする。
　計測日時毎に、電気使用量データと、ベースラインと、を比較し、ベースラインが電気使用量データより大きい場合は、該当するベースラインを電気使用量データに補正するとともに、空調等電気使用量を０に補正する。
　ベースラインと、ベース電気使用量と、の差を照明等電気使用量とする。
　図８の実施例は、計測日時毎で、計測時間帯フラグと、気温と、電気使用量と、稼働／非稼働と、空調の使用有無と、ベースラインと、空調等、照明等、ベースの用途別電気使用量をまとめた一例である。例えば、ベースライン（非稼働：１０００）は、計測時間帯１０００での非稼働のベースラインを示し、ベースライン（非稼働：３０）’は、計測時間帯３０での非稼働ベースラインを補正した値を示し、電気使用量［１１０］は、電気使用量が１１０ｋＷｈであることを示している。
　稼働／非稼働フラグ、空調等、照明等、ベースの用途別電気使用量を、記憶部２３の分析結果データベース２３５に記録する。
　電気使用量予測パッケージ２４４は、電気使用量データと、気象データと、を用いて、稼働／非稼働別電気使用量予測モデルを作成し、稼働／非稼働別電気使用量予測モデルに気象予報値を入力し、電気使用量予測値を計算するとともに、電気使用量予測値を用いて、空調等、照明等、ベースの用途別電気使用量を予測する。電気使用量予測値の確率分布を作成して、電気使用量予測値の発生確率を計算する。
　用途別電気使用量推計パッケージ２４３の計算結果を用いて、稼働／非稼働フラグを修正する。計測時間帯別稼働／非稼働別ベースラインの中で非稼働である計測日時の集合である計測時間帯別非稼働ベースラインの最大値より、電気使用量が小さい計測日時は、非稼働にして、稼働／非稼働フラグを修正する。
　計測日時の曜日、祝日、時間帯等を入力、前記修正した稼働／非稼働フラグを出力とした稼働回帰モデルを作成する。回帰モデルの作成には、ロジスティック回帰分析やサポートベクトルマシン分析等を用いることが考えられる。
　予測を行う期間の稼働／非稼働を予測するために、稼働回帰モデルに、予測する日時の曜日、祝日、時間帯等を入力して、稼働／非稼働予測フラグを計算する。
　計測時間帯別に、修正した稼働／非稼働別に、過去の気象データを入力、電気使用量データを出力とした正則化項を含めた多項式回帰分析又は、カーネル回帰分析を行い、計測時間帯別稼働／非稼働別電気使用量予測モデルを作成する。
　計測時間帯別稼働／非稼働別電気使用量予測モデルに、気象予報値を入力して、電気使用量予測値と、前記電気使用量予測値の確率分布を計算する。
　計測時間帯別稼働／非稼働別電気使用量予測モデルと、稼働回帰モデルと、を保存しておき、リアルタイムでの電気使用量推計に活用する。電気使用量予測モデルと、稼働回帰モデルと、は１か月毎等定期的に更新する。
　電気使用量予測値と、用途別電気使用量計算モジュール２４３５で計算された稼働計測時間帯別ベースラインと、ベース電気使用量と、を用いて、用途別電気使用量計算モジュール２４３５と同等の処理を行い、空調等電気使用量予測値、照明等電気使用量予測値、ベース電気使用量予測値を計算する。
　図９は、過去の電気使用量データと、過去の空調等、照明等、ベースの用途別電気使用量と、電気使用量予測値と、空調等、照明等、ベースの用途別電気使用量予測値と、電気使用量予測値の確率分布と、を一括でグラフ表示した図の一例である。
　図９は、横軸は日時、縦軸は電気使用量を示し、過去の電気使用量データ２４４ｉと、電気使用量予測値２４４ｊを線グラフで示し、過去のベース電気使用量２４４ｃと、照明等電気使用量２４４ｄと、空調等電気使用量２４４ｅと、の用途別電気使用量を積み上げ棒グラフで示し、ベース電気使用量の予測値２４４ｆと、照明等電気使用量の予測値２４４ｇと、空調等電気使用量の予測値２４４ｈと、の用途別電気使用量の予測値を積み上げ棒グラフで示す。異常値が発生する確率を、前もって設定しておき、電気使用量予測値の確率分布から、閾値が、前記異常値が発生する確率の１／２となる下限の値を予測下限値２４４ａとして、閾値が、前記異常値が発生する確率の１／２となる上限の値を予測上限値２４４ｂとして面グラフで示す。例えば、異常値が発生する確率を５％と設定した場合、予測下限値２４４ａと、予測上限値２４４ｂと、の範囲は、確率分布として９５％の確率で発生する予測値の範囲となる。このグラフ表示方法により、電気使用量の異常は発生したか、今後の電気使用量がどのように推移するのか等が一つのグラフで確認でき、省エネ対策に役立てることができる。
　電気使用量予測値と、空調等、照明等、ベースの用途別電気使用量予測値と、電気使用量予測値の確率分布と、を記憶部２３の分析結果データベース２３５に記録する。
　省エネ分析パッケージ２４５は、用途別電気使用量推計パッケージ２４３の計算結果を用いて、省エネにつながる分析を行う。
　省エネ分析パッケージ２４５は、例えば、空調設定温度モジュール２４５１やデマンド分析モジュール２４５２、時間外分析モジュール２４５３を含む。
　空調設定温度モジュール２４５１は、空調設定温度を変更した場合の省エネ量を計算する。事前に空調設定温度の変化量を設定して、省エネ計算用気温を計算する。省エネ計算用気温は、計測日時毎に計算する。
　図７に示した、稼働時の電気使用量データの冷房範囲及び暖房範囲の境界線２４３８ｅを、対象計測時間帯における稼働時の気温閾値、非稼働時の電気使用量データの冷房範囲及び暖房範囲の境界線２４３８ｆを、対象計測時間帯における非稼働時の気温閾値とする。計測日時における、稼働／非稼働及び計測時間帯が一致する気温閾値を用いて、対象計測日時の省エネ計算用気温を計算する。前記気温閾値より、気温が高い場合は、気温から空調設定温度の変化量を引いた値とする。ただし、引いた値が気温閾値より小さくなる場合は、省エネ計算用気温を気温閾値と同じ値にする。気温が気温閾値以下の場合は、気温から前記空調設定温度の変化量を加えた値とする。ただし、加えた値が気温閾値より大きくなる場合は、省エネ計算用気温を気温閾値と同じ値にする。
　用途別電気使用量計算モジュール２４３５に、気温データを省エネ計算用気温に置換した気象データを入力して、省エネ計算用電気使用量を推計する。計測日時毎に、過去データから計算された電気使用量推計値と、省エネ計算用電気使用量と、の差を計算して、設定変更による省エネ量とする。ただし、電気使用量推計値と、省エネ計算用電気使用量と、の差が負になる場合は、該当する計測日時の設定変更による省エネ量を０とする。
　デマンド分析モジュール２４５２は、合理的な電力デマンド値の目標値を計算して、合理的な電力デマンド監視を促す。
　事前に、異常値の発生確率を定めておく。
　電気使用量データの３０分間隔の合計値と、気象データの３０分間隔の平均値と、を用途別電気使用量推計パッケージ２４３に入力する。
　電気使用量データの３０分間隔の合計値の最大値を２倍した値を、現状の電力デマンド値とする。
　用途別電気使用量推計パッケージ２４３で導出した計測時間帯別稼働／非稼働別電気使用量推計モデルから計算される電気使用量推計値の最大値を２倍した値を、達成可能なデマンド値とする。
　計測時間帯別稼働／非稼働別電気使用量推計モデルを作成する際に、推計値の確率分布を計算する。前記確率分布から、前記異常値の発生確率が閾値となる推計値を計算し、その値を２倍した値を達成容易なデマンド値とする。
　空調設定温度モジュール２４５１に入力し、得られた省エネ計算用電気使用量推計値の最大値を２倍した値を、設定温度変更によるデマンド値とする。
　達成可能な電力デマンド値、達成容易な電力デマンド値、設定温度変更によるデマンド値は、それぞれ現状の電力デマンド値より大きい場合は、現状の電力デマンド値に補正する。
　図１０は、横軸を時間、縦軸を電力デマンド値とし、電気使用量と、ベース電気使用量２４ｋ、照明等電気使用量２４ｌ、空調等電気使用量２４ｍの用途別電気使用量を組み合わせたデータと、を電気使用量の降順に並べ替えて、積み上げ棒グラフで表示した図の一例である。
　電力デマンド値２４ｎは、現状の電力デマンド値と、達成容易な電力デマンド値と、達成可能な電力デマンド値と、設定温度変更による電力デマンド値をそれぞれ示す。このグラフ表示方法により、電力デマンド値の合理的な目標値や最大電力デマンド値が発生している際の、用途別電気使用量の内訳等をわかりやすく表示できる。
　時間外分析モジュール２４５３は、残業等の所定時間外労働による追加的な電気使用量を計算し、時間外労働の削減による省エネを促す。
　用途別電気使用量推計パッケージ２４３を用いて稼働／非稼働フラグと、計測時間帯別稼働／非稼働別電気使用量推計モデルと、計測時間帯別稼働／非稼働別ベースラインと、を計算する。
　用途別電気使用量推計パッケージ２４３の計算結果を用いて、稼働／非稼働フラグを修正する。計測時間帯別稼働／非稼働別ベースラインの中で非稼働である計測日時の集合である計測時間帯別非稼働ベースラインの最大値より、電気使用量が小さい計測日時は、非稼働にして、稼働／非稼働フラグを修正する。
　計測日時の曜日、祝日、時間帯等を入力、前記修正した稼働／非稼働フラグを出力とした稼働回帰モデルを作成する。回帰モデルの作成には、ロジスティック回帰分析やサポートベクトルマシン分析等を用いることが考えられる。
　稼働回帰モデルに、計測日時の曜日、祝日、時間帯等を入力して、建物の定期的な稼働／非稼働を計算する。
　定期的な非稼働の計測日時において、修正した稼働／非稼働フラグが稼働である場合に、所定時間外労働としてフラグを立てる。
　計測時間帯別非稼働電気使用量推計モデルを用いて、所定時間外労働の計測日時における非稼働電気使用量を推計する。
　所定時間外労働フラグが立っている計測日時における電気使用量と、所定時間外労働の計測日時における非稼働電気使用量と、の差を所定時間外労働による追加的な電気使用量として計算する。
　省エネ分析パッケージ２４５で計算された結果は、記憶部２３の分析結果データベース２３５に記録する。
　以上に説明したように、本実施形態によって、特別な専門的知識が無くても、建物の電気使用量と、建物住所情報という容易に入手可能な情報のみで、多数のオフィス、工場等の建物を対象に省エネ診断が可能となる。
　本実施形態で必要となる、建物の電気使用量と住所情報は、電力会社等が顧客を対象として容易に入手可能な情報であり、詳細な省エネ診断を定期的に送付することや、リアルタイムで診断結果を、ネットワークを通して確認するサービスにも利用できる。
　また、オフィス、工場等の多数の顧客の建物を対象に省エネ診断を行うことで、省エネ効果の高い建物を抽出し、効率の高い省エネ対策を推進できる。
　ここでは好ましい形態の一例を示しており、記載した内容以外においても実施は可能である。また、ここで紹介した式、方法等については代表例を示したが、これらに限定するものではなく、他の公式についても代用可能である。 Hereinafter, specific embodiments of the present invention will be described with reference to the drawings.
The data measurement period used in the present embodiment can be analyzed if it is data of 6 months or more.
FIG. 2 is a block diagram illustrating an example of the configuration of the energy saving diagnosis system 10 according to the present embodiment.
The energy saving diagnosis system 10 according to the present embodiment includes a client-side machine 3000, an external weather information database 4000, an external location information database 5000, an external communication network 6000, and a server 1000 as an analysis system.
The client side machine 3000 sends the electricity usage data 3002 and the building address data 3001 described later to the server 1000 as the analysis system through the external communication network 6000, and sends the analysis result of the analysis system to the client side machine 3000. . Meteorological observation facility location information 4001, meteorological observation data 4002, and meteorological forecast data 4003, which will be described later, are periodically sent from the external meteorological information database 4000 to the server 1000 as an analysis system through the external communication network 6000. From the server 1000 in which the analysis system is introduced, the building address data 3001 is sent to the external location information database 5000 through the external communication network 6000, and latitude, longitude, and altitude data necessary for analysis are obtained.
The building address data 3001 is information indicating a building address or zip code information.
The electricity usage data 3002 is electricity usage data measured at regular intervals (for example, at 30-minute intervals) from a smart meter or the like.
The weather observation facility location information data 4001 is latitude, longitude, and altitude data of the weather observation facility.
The meteorological observation data 4002 is data observed at a weather station such as temperature, humidity, sunshine duration, wind speed, wind direction, and precipitation.
The weather forecast data 4003 is data predicted at a weather station such as temperature, humidity, sunshine duration, wind speed, wind direction, and precipitation.
The position information database 5000 is latitude, longitude, altitude data, and the like.
The communication network 6000 is an Internet network, an intranet network, or the like.
The server 1000 includes, for example, the communication device 1 and the processing device 2. The communication device 1 is an arbitrary device that is communicably connected to the external communication network 6000.
The processing device 2 executes various control processes in the server 1000. The processing device 2 includes one or a plurality of servers including a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), an I / O interface (Input-Output Interface), and the like. Functions of each part are realized by causing a computer to execute a predetermined program.
FIG. 3 is a block diagram illustrating an example of the configuration of the processing apparatus 2 according to the present embodiment.
The processing apparatus 2 includes, for example, a communication processing unit 21, an input unit 22, a storage unit 23, an analysis unit 24, and an output unit 25.
The communication processing unit 21 controls the communication device 1 and acquires weather observation facility position information 4001, weather observation data 4002, weather forecast data 4003, building address data 3001, and electricity usage data 3002 via the external communication network 6000. To do.
The input unit 22 acquires the weather observation facility location information 4001, weather observation data 4002, weather forecast data 4003, building address data 3001, and electricity usage data 3002 from the communication processing unit 21, and the weather observation facility in the storage unit 23. The database 231, the weather database 232, the building address database 233, and the electricity usage database 234 are stored in corresponding databases.
The analysis unit 24 acquires necessary data from the input unit 22 and the storage unit 23. Further, the analysis unit 24 is connected to the external position information database 5000 through the communication processing unit 21 through a Web API (Web Application Programming Interface) or the like, transmits building address data 3001, and building position information such as latitude, longitude, and altitude. To get. The configuration of the analysis unit 24 will be described in detail later. The results analyzed by the analysis unit 24 are stored in the analysis result database 235 in the storage unit 23.
The output unit 25 acquires necessary data from the analysis unit 24, and outputs the analysis result on the client side machine 3000 through the communication processing unit 21.
It is also assumed that the building address data 3001 and the electricity usage data 3002 do not require input from the outside and are already stored in the database.
The input unit 22, the analysis unit 24, and the output unit 25 are assumed to be physically separate servers or the same server.
The output of the analysis result may be displayed on the client side machine 3000 through the output unit 25 or may be used for analysis on the analysis result database 235.
FIG. 1 is a block diagram illustrating an example of the configuration of the analysis unit 24 of the processing apparatus 2 according to the present embodiment.
The analysis unit 24 includes, for example, a weather data acquisition package 241, a data preprocessing package 242, a usage-specific electricity usage estimation package 243, an electricity usage prediction package 244, and an energy saving analysis package 245.
A package is a set of program codes roughly classified by function, and a module is a set of program codes roughly classified by analysis content.
The weather data acquisition package 241 includes, for example, a position information acquisition module 2411 and a weather data extraction module 2412.
The location information acquisition module 2411 transmits the building address data 3001 of the building to be analyzed from the building address database 233 of the storage unit 23 through the communication processing unit 21 to the external location information database 5000 through the Web API or the like, and the latitude of the building, Acquire location information such as longitude and altitude. Using the acquired building position information, for example, a weather observation facility within the range of latitude ± 1 degree and longitude ± 1 degree of the building position information is extracted from the weather observation facility database 231 of the storage unit 23. The distance between the two points of the building and the extracted weather observation facility is calculated, and the weather observation facility closest to the building is selected.
The meteorological data extraction module 2412 executes the analysis of the meteorological data of the meteorological observation facility located closest to the building to be analyzed from the meteorological database 232 of the storage unit 23 in the electricity usage database 234 of the storage unit 23. To extract the electricity usage data and the same period. Weather forecast data is also extracted.
The data preprocessing package 242 acquires electricity usage data, weather data, and weather forecast data.
The data preprocessing package 242 matches the electricity usage data, weather data, and weather forecast data at the same measurement interval. Basically, up-sampling or down-sampling is performed on weather data and weather forecast data in accordance with the measurement interval of electricity usage data. Upsampling uses linear interpolation, spline interpolation, or the like. Each data after the interpolation is analyzed using an abnormal value detection method, and if an abnormal value is found, it is deleted and treated as a missing data. If each data is missing for a short time, the data is interpolated by linear interpolation or the like. If it is a long-term deficit, leave it deficient.
As shown in FIG. 4, the data preprocessing package 242 combines the electricity usage data 242c and the weather data 242d using the measurement date 242a as a key, and creates a measurement time zone flag 242b from the measurement date 242a. To create analysis data. The measurement time zone flag 242b is calculated as, for example, the value of time of measurement date × 100 + the value of minute of measurement date.
Electricity usage estimation package 243 by application
It is a package that estimates electricity usage by application from analysis data. The usage-specific electricity usage estimation package 243 includes, for example, a discrimination module 2431, a model type estimation module 2433, and a usage-specific electricity usage calculation module 2435.
The determination module 2431 is a module that determines whether the building is operating / not operating at the measurement date and time.
About analysis data, operation / non-operation is discriminated from a set of a pair of electricity usage data and weather data. As a discrimination method, a method using a K-means algorithm, a method using a conditional mixture model, and the like can be considered. Here, a method using a support vector machine will be described.
The method using the support vector machine requires training data for operation / non-operation. For training data for operation / non-operation, the range from the minimum value to the maximum value of the temperature of the analysis data is divided into 8 or 10 for example, and several points are counted from the maximum in the electricity usage data within the range. The training data is activated, and several points from the minimum are extracted as non-operational training data. A discriminant model is created from the training data using a support vector machine, and the remaining analysis data is discriminated as active / non-operating using the discriminant model.
FIG. 5 shows an example of a building operation / non-operation determination method for the determination module 2431.
In FIG. 5, the horizontal axis indicates the air temperature, the vertical axis indicates the electricity usage, the analysis data of the pair of electricity usage data and weather data is the white circle (2431a), the operation training data is the black circle (2431b), and the non-operation training Data is indicated by a cross (2431c), and a dividing line when the range from the minimum value to the maximum value is divided into eight is indicated by a dotted line (2431d).
The model type estimation module 2433 creates an electricity usage amount estimation model for each operation time / non-operation time to estimate the electricity usage by using the data for each measurement time zone extracted from the analysis data for each measurement time zone. Yes, this will be described with reference to the flowchart of FIG.
The estimation of the amount of electricity used is started from the measurement time zone data (S201).
The processing from S202 to S209 is repeated for each measurement time period.
The data classified by measurement time zone is discriminated as active and non-operating using the discrimination module 2431 (S203 to S204).
Training data when a support vector machine is used for the discrimination module 2431 includes a method of dividing the temperature into a plurality of parts and operating the upper few points of the electricity usage data and disabling the lower several points, and the model type estimation module 2433. Before the operation, the period operation / non-operation flag extracted in the measurement time zone among the operation / non-operation flags created by the discrimination module 2341 using the weather data and the electricity usage data is used as training data. The method to be used is considered.
Extracts data by measurement time zone for each operation / non-operation, inputs meteorological data, outputs electricity usage data, and performs polynomial regression analysis including regularization terms or kernel regression analysis for each measurement time zone An electricity usage estimation model for each operation / non-operation is created (S205 to S208).
A probability distribution of predicted electric usage values is created for the operating / non-operating flag and the weather data, using the electric usage amount estimation model for each operating time / non-operating time. The probability distribution indicates an output generation probability with respect to a certain input.
For each measurement time period, the model type estimation module 2433 is executed, and the obtained results are combined (2432 to 2434).
The usage-specific electricity usage calculation module 2435 will be described with reference to FIGS.
The electricity usage calculation module 2435 for each application receives the electricity usage estimation model for each operation time / non-operation time and the operation / non-operation flag, and uses electricity such as air conditioning for each operation time / non-operation time. The amount, the electricity usage such as lighting, and the base electricity usage are calculated.
Within the range of the meteorological data of the data by measurement time zone from the electricity usage estimation model by operation time / non-operation, the operation / The minimum value of estimated electricity usage by non-operation is set as the baseline by operation / non-operation by measurement time period. The temperature from which the minimum value is derived is defined as a temperature threshold value, a temperature range higher than the temperature threshold value is defined as a cooling range, and a temperature range equal to or lower than the temperature threshold value is defined as a heating range.
FIG. 7 is a diagram illustrating an example of a method for determining the cooling range and the heating range of the electric usage amount.
In FIG. 7, the horizontal axis indicates the temperature, the vertical axis indicates the electricity usage, the electricity usage data during operation is a white circle (2438a), the electricity usage data during non-operation is a cross (2438b), and the electricity during operation The boundary line between the cooling range and the heating range of the usage amount data is a dotted line 2438e, and the boundary line of the cooling range and the heating range of the non-operating electricity usage data is a dotted line 2438f, which is the minimum value of the estimated electricity usage amount during the operation. The base line is indicated by a dotted line 2438c, and the base line which is the minimum value of the estimated electricity consumption when not in operation is indicated by a dotted line 2438d.
There may be cases where the cooling range and the heating range are different for operation and non-operation. When the air temperature data and the electricity usage data are shown to have a significant positive correlation using the data for each operating time / non-operation in the measurement range of the cooling range, it is considered that there is cooling and the electricity usage The difference between the data and the baseline for each operating time / non-operating time is the amount of electricity used for air conditioning. If it is not significant, the electricity usage amount such as air conditioning is set to 0, and the baseline of the corresponding measurement date is corrected to the electricity usage amount of the relevant measurement date. Similarly, if it is shown that there is a significant negative correlation between the temperature data and the electricity usage data using the operating range data for each heating time range, the electricity usage data and The difference between the operating time and the non-operating baseline for each measurement time zone is the amount of electricity used for air conditioning. If it is not significant, the electricity usage amount such as air conditioning is set to 0, and the baseline of the corresponding measurement date is corrected to the electricity usage amount of the relevant measurement date.
Of all the baselines by operating time / non-operating by measurement time zone before correction, the minimum baseline value is used as the base electricity consumption.
At each measurement date and time, the electricity usage data is compared with the baseline. If the baseline is larger than the electricity usage data, the corresponding baseline is corrected to the electricity usage data and the electricity usage such as air conditioning. Is corrected to 0.
The difference between the baseline and the base electricity usage is the electricity usage for lighting.
In the embodiment of FIG. 8, the measurement time zone flag, temperature, electricity usage, operation / non-operation, use / non-use of air conditioning, baseline, air conditioning, etc. It is an example which put together the electric consumption according to a use. For example, the baseline (non-operating: 1000) indicates the non-operating baseline in the measurement time zone 1000, and the baseline (non-operating: 30) 'is the non-operating baseline in the measuring time zone 30 corrected. The value indicates that the electricity usage [110] indicates that the electricity usage is 110 kWh.
The base usage-specific electricity usage, such as operating / non-operating flag, air conditioning, etc., is recorded in the analysis result database 235 of the storage unit 23.
The electricity usage prediction package 244 creates an electricity usage prediction model for each operation / non-operation by using the electricity usage data and the weather data, and the weather prediction value for the electricity usage prediction model for each operation / non-operation. Is used to calculate a predicted electricity usage amount, and the electricity usage amount for each base application such as air conditioning or lighting is predicted using the electricity usage prediction value. A probability distribution of predicted electricity usage is created, and the occurrence probability of the predicted electricity usage is calculated.
The operating / non-operating flag is corrected using the calculation result of the usage-specific electricity usage estimation package 243. The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are not in operation in the baseline by operation time and non-operation time. , Correct the operation / non-operation flag.
An operation regression model is created by inputting the day of the week, holiday and time zone of the measurement date and time, and outputting the modified operation / non-operation flag. Logistic regression analysis, support vector machine analysis, or the like can be used to create a regression model.
In order to predict the operation / non-operation in the period for which the prediction is performed, the operation / non-operation prediction flag is calculated by inputting the day of the week, the holiday, the time zone, etc. of the prediction date and time into the operation regression model.
Operate by measurement time period by performing polynomial regression analysis or kernel regression analysis including regularized terms with past meteorological data input and electricity usage data output for each modified operation / non-operation. / Create a forecast model for electricity usage by non-operation.
A weather forecast value is input to the electricity usage amount prediction model for each operation time / non-operation time, and the electricity usage amount prediction value and the probability distribution of the electricity usage amount prediction value are calculated.
The electricity usage prediction model for each operation time / non-operation and the operation regression model are stored and used for estimation of electricity usage in real time. The electricity usage prediction model and the operation regression model are updated periodically such as every month.
Processing equivalent to the usage-specific electricity usage calculation module 2435 using the predicted electricity usage, the baseline for each operation measurement time period calculated by the usage-specific electricity usage calculation module 2435, and the base electricity usage To calculate a predicted electricity usage amount such as air conditioning, a predicted electricity usage amount such as lighting, and a predicted base electricity usage amount.
FIG. 9 shows past electricity usage data, past air conditioning, etc., lighting, etc., base usage of electricity, predicted usage, and air conditioning, lighting, etc., base usage, electricity usage prediction It is an example of the figure which displayed the value and the probability distribution of the electricity usage predicted value collectively as a graph.
In FIG. 9, the horizontal axis indicates the date and time, the vertical axis indicates the electricity usage, the past electricity usage data 244i and the electricity usage predicted value 244j are shown in a line graph, the past base electricity usage 244c, the lighting, etc. The electric usage amount 244d and the electric usage amount 244e for air conditioning, etc. are shown in a stacked bar graph, the predicted value 244f for the base electric usage amount, the predicted value 244g for the electric usage amount for lighting, etc. The predicted value of the usage amount 244h and the predicted value of the electricity usage amount by use are shown in a stacked bar graph. The probability that an abnormal value occurs is set in advance, and from the probability distribution of the electricity usage predicted value, a lower limit value at which the threshold value is ½ of the probability that the abnormal value occurs is set as a predicted lower limit value 244a. An upper limit value at which the threshold value is ½ of the probability of occurrence of the abnormal value is shown as a predicted upper limit value 244b in a plane graph. For example, when the probability of occurrence of an abnormal value is set to 5%, the range between the prediction lower limit value 244a and the prediction upper limit value 244b is a range of prediction values that occur with a probability of 95% as a probability distribution. With this graph display method, it is possible to confirm whether an abnormality in the amount of electricity used has occurred or how the amount of electricity used in the future will change with a single graph, which can be used for energy saving measures.
The predicted electric usage amount, the predicted electric usage amount for each base such as air conditioning and lighting, and the probability distribution of the predicted electric usage amount are recorded in the analysis result database 235 of the storage unit 23.
The energy saving analysis package 245 performs an analysis that leads to energy saving using the calculation result of the usage-specific electricity usage estimation package 243.
The energy saving analysis package 245 includes, for example, an air conditioning set temperature module 2451, a demand analysis module 2452, and an after-hours analysis module 2453.
The air conditioning set temperature module 2451 calculates an energy saving amount when the air conditioning set temperature is changed. Set the amount of change in air conditioning set temperature in advance and calculate the temperature for energy saving calculation. The temperature for energy saving calculation is calculated for each measurement date.
The boundary line 2438e of the cooling range and heating range of the electricity usage data during operation shown in FIG. 7 is the temperature threshold during operation in the target measurement time zone, the cooling range and heating range of the electricity usage data during non-operation. The boundary line 2438f is set as a temperature threshold value during non-operation in the target measurement time zone. The temperature for energy saving calculation at the target measurement date / time is calculated using the temperature threshold value at which the operation / non-operation and the measurement time zone match at the measurement date / time. When the air temperature is higher than the air temperature threshold, a value obtained by subtracting the change amount of the air conditioning set temperature from the air temperature is used. However, if the subtracted value is smaller than the temperature threshold, the temperature for energy saving calculation is set to the same value as the temperature threshold. When the air temperature is equal to or lower than the air temperature threshold, a value obtained by adding the change amount of the air conditioning set temperature to the air temperature is used. However, when the added value is larger than the temperature threshold, the temperature for energy saving calculation is set to the same value as the temperature threshold.
Meteorological data obtained by replacing the temperature data with the temperature for energy saving calculation is input to the usage-specific electricity usage calculation module 2435 to estimate the energy usage for electricity saving calculation. For each measurement date and time, the difference between the estimated value of electricity usage calculated from past data and the electricity usage for energy saving calculation is calculated to obtain the energy saving amount by changing the setting. However, if the difference between the estimated value of electricity usage and the electricity usage for energy saving calculation is negative, the energy saving amount by changing the setting of the corresponding measurement date is set to zero.
The demand analysis module 2452 calculates a target value for a reasonable power demand value to facilitate reasonable power demand monitoring.
The occurrence probability of abnormal values is determined in advance.
The total value of the electricity usage data at the 30-minute interval and the average value of the weather data at the 30-minute interval are input to the usage-specific electricity usage estimation package 243.
A value obtained by doubling the maximum value of the total value of the electricity usage data every 30 minutes is set as the current power demand value.
Achievable demand value is obtained by doubling the maximum value of the estimated electricity usage calculated from the electricity usage estimation model by operation time / non-operation time derived by the electricity usage estimation package 243 by application. To do.
The probability distribution of the estimated value is calculated when the electricity usage amount estimation model for each operation time period is created. From the probability distribution, an estimated value having the occurrence probability of the abnormal value as a threshold value is calculated, and a value obtained by doubling the value is set as a demand value that can be easily achieved.
A value that is input to the air conditioning set temperature module 2451 and that is obtained by doubling the maximum value of the estimated electricity use amount for energy saving calculation is set as a demand value by changing the set temperature.
If the achievable power demand value, the easily achievable power demand value, and the demand value by changing the set temperature are larger than the current power demand value, they are corrected to the current power demand value.
In FIG. 10, the horizontal axis represents time, the vertical axis represents power demand value, and the amount of electricity used is combined with the amount of electricity used for each application, such as 24k for base electricity, 24l for electricity such as lighting, and 24m for electricity such as air conditioning. It is an example of the figure which rearranged data and displayed in the descending order of the amount of electricity used, and was displayed by the stacked bar graph.
The power demand value 24n indicates a current power demand value, an easily achieved power demand value, an achievable power demand value, and a power demand value by changing the set temperature. With this graph display method, it is possible to easily display the breakdown of the amount of electricity used by application when a reasonable target value of the power demand value or the maximum power demand value is generated.
The overtime analysis module 2453 calculates an additional amount of electricity used for overtime work such as overtime, and promotes energy saving by reducing overtime work.
Calculate the operating / non-operating flag, the operating / non-operating electricity usage estimation model by measurement time zone, and the operating / non-operating baseline by measurement time zone using the usage-specific electricity usage estimation package 243 .
The operating / non-operating flag is corrected using the calculation result of the usage-specific electricity usage estimation package 243. The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are not in operation in the baseline by operation time and non-operation time. , Correct the operation / non-operation flag.
An operation regression model is created by inputting the day of the week, holiday and time zone of the measurement date and time, and outputting the modified operation / non-operation flag. Logistic regression analysis, support vector machine analysis, or the like can be used to create a regression model.
Input the day of the week, holiday, time zone, etc. of the measurement date and time into the operation regression model to calculate the regular operation / non-operation of the building.
If the corrected working / non-working flag is working at the regular non-working measurement date and time, a flag is set as overtime work.
Non-operating electricity usage at the measurement date and time of overtime work is estimated using the non-operating electricity usage estimation model by measurement time zone.
The difference between the amount of electricity used at the measurement date and time when the overtime work flag is set and the amount of non-operating electricity used at the measurement date and time of overtime work is calculated as an additional amount of electricity used by overtime work.
The result calculated by the energy saving analysis package 245 is recorded in the analysis result database 235 of the storage unit 23.
As described above, according to the present embodiment, a large number of buildings such as offices and factories can be obtained by using only easily available information such as the amount of electricity used in the building and the building address information without special technical knowledge. Energy-saving diagnosis is possible for
The electricity usage and address information of the building required in this embodiment is information that can be easily obtained by the power company etc. for customers, sending detailed energy-saving diagnosis regularly, or in real time It can also be used for services that check the diagnosis results via the network.
Also, by conducting energy-saving diagnosis for buildings of many customers such as offices and factories, it is possible to extract buildings with high energy-saving effects and promote highly efficient energy-saving measures.
Here, an example of a preferable form is shown, and implementation is possible in addition to the contents described. Moreover, although the representative example was shown about the formula, method, etc. which were introduced here, it is not limited to these, Other formulas can be substituted.
　２　　　　　処理装置
　２１　　　　通信処理部
　２３　　　　記憶部
　２４　　　　分析部
　２３１　　　気象観測施設データベース
　２３２　　　気象データベース
　２３３　　　建物住所データベース
　２３４　　　電気使用量データベース
　２３５　　　分析結果データベース
　２４１　　　気象データ取得パッケージ
　２４２　　　データ前処理パッケージ
　２４３　　　用途別電気使用量推計パッケージ
　２４４　　　電気使用量予測パッケージ
　２４５　　　省エネ分析パッケージ
　２４３１　　判別モジュール
　２４３３　　モデル式推計モジュール
　２４３５　　用途別電気使用量計算モジュール
　２４５１　　空調設定温度モジュール
　２４５２　　デマンド分析モジュール
　２４５３　　時間外分析モジュール
　３０００　　クライアント側のマシン
　３００１　　建物住所データ
　３００２　　電気使用量データ
　４０００　　気象情報データベース
　４００１　　気象観測施設位置情報
　４００２　　気象観測データ
　４００３　　気象予報データ
　５０００　　位置情報データベース 2 Processing Device 21 Communication Processing Unit 23 Storage Unit 24 Analysis Unit 231 Weather Observation Facility Database 232 Weather Database 233 Building Address Database 234 Electricity Usage Database 235 Analysis Result Database 241 Weather Data Acquisition Package 242 Data Preprocessing Package 243 Electricity Usage by Use Estimation package 244 Electricity usage forecast package 245 Energy saving analysis package 2431 Discrimination module 2433 Model type estimation module 2435 Electricity usage calculation module 2451 by use Air conditioning setting temperature module 2452 Demand analysis module 2453 After-hours analysis module 3000 Client side machine 3001 Building address Data 3002 Electricity usage data 4000 Weather information database 4001 weather observation facility location information 4002 weather observation data 4003 weather forecast data 5000 position information database

Claims (6)

  1. 　コンピュータに、
    　外部から得られる定期的な周期で計測された建物の電気使用量データと前記建物周辺の気象データを用いて、計測日時における前記建物の稼働／非稼働と、空調等、照明等、ベースの用途別電気使用量と、を計算させるプログラムであって、
    　判別モジュールと、モデル式推計モジュールと、用途別電気使用量計算モジュールと、から構成され、
    　前記判別モジュールは、計測時間帯別に、前記電気使用量データと前記気象データを抽出した計測時間帯別データを用いて、計測日時における稼働と、非稼働と、を判別する判別モデルを作成し、前記計測時間帯における、前記対象計測日時における稼働と、非稼働と、の判別を実施して、
    　前記モデル式推計モジュールは、前記判別モジュールで判別した稼働／非稼働と、前記計測時間帯別データと、を用いて、前記計測時間帯別に、稼働／非稼働別に、前記計測時間帯別データの気象データを入力、電気使用量データを出力とした回帰分析を行い、計測時間帯別稼働／非稼働別電気使用量推計モデルを作成して、
    　前記用途別電気使用量計算モジュールは、前記計測時間帯別稼働／非稼働別電気使用量推計モデルから、前記計測時間帯別データの気象データの外気温又は、外気温と湿度との、上限値から下限値の範囲内において、前記計測時間帯別稼働／非稼働別電気使用量推計モデルから計算された電気使用量推計値の最小値を計測時間帯別稼働／非稼働別ベースラインとして計算して、
    　前記計測時間帯において、稼働／非稼働別に、前記電気使用量データと、前記計測時間帯別稼働／非稼働別ベースラインと、の差を、空調等電気使用量として計算して、
    　全ての前記計測時間帯別稼働／非稼働別ベースラインの中で、最小の値をベース電気使用量として計算して、
    　前記計測時間帯別稼働／非稼働別ベースラインと、前記ベース電気使用量と、の差をそれぞれ計算した値を、照明等電気使用量として計算して、
    　前記計測時間帯別稼働／非稼働別電気使用量推計モデルと、稼働／非稼働フラグと、空調等、照明等、ベースの用途別電気使用量と、を計算させるプログラム。     On the computer,
    Using the building's electricity usage data measured at regular intervals from the outside and the weather data around the building, the use of the building at the measurement date and time, air conditioning, lighting, etc. A program for calculating the amount of electricity used separately,
    It consists of a discrimination module, a model type estimation module, and an electricity usage calculation module for each application.
    The discriminating module creates a discriminant model for discriminating between operation and non-operation at the measurement date and time by using the data for each measurement time zone obtained by extracting the electricity usage data and the weather data for each measurement time zone, In the measurement time zone, to determine the operation and non-operation at the target measurement date and time,
    The model type estimation module uses the operation / non-operation determined by the determination module and the data for each measurement time zone, and uses the data for each measurement time zone for each measurement time zone, for each operation / non-operation. Perform regression analysis with weather data as input and electricity usage data as output, and create electricity usage estimation model by operation time / non-operation time,
    The usage-specific electricity usage calculation module is based on the measurement time zone operation / non-operational electricity usage estimation model, the outside temperature of the weather data of the measurement time zone data, or the upper limit value of the outside temperature and humidity Within the range of the lower limit value, calculate the minimum value of the estimated electricity usage calculated from the above-mentioned operating / non-operating electricity usage estimation model by measurement time period as the baseline by operating time / non-operational period by measurement time period. And
    In the measurement time zone, for each operation / non-operation, the difference between the electricity usage data and the baseline for each measurement time zone operation / non-operation is calculated as the amount of electricity used for air conditioning,
    Calculate the minimum value as the base electricity consumption in all the above-mentioned baselines by operation time / non-operation time,
    Calculate the value calculated as the difference between the baseline for each operation time / non-operation for each measurement time period and the amount of base electricity used as the amount of electricity used for lighting, etc.
    A program for calculating the electricity usage estimation model for each operation time / non-operation, the operation / non-operation flag, and the electric usage for each base such as air conditioning and lighting.
  2. 　コンピューターに、
    　外部から入手する気象予報値と、前記気象データと、前記電気使用量データと、を用いて、電気使用量予測値と、前記電気使用量予測値の発生確率分布と、空調等、照明等、ベースの用途別電気使用量予測値と、を計算させるプログラムであって、
    　請求項１を用いて前記稼働／非稼働フラグと、前記計測時間帯別稼働／非稼働別ベースラインと、前記ベース電気使用量と、を計算して、
    　前記計測時間帯別稼働／非稼働別ベースラインの中で非稼働である計測日時の集合である前記計測時間帯別非稼働ベースラインの最大値より、前記電気使用量が小さい計測日時は、非稼働にして、前記稼働／非稼働フラグを修正して、
    　前記計測日時の曜日、祝日、時間帯等を入力、前記修正した稼働／非稼働フラグを出力とした稼働回帰モデルを作成して、
    　前記稼働回帰モデルに、予測する日時の曜日、祝日、時間帯等を入力して、稼働／非稼働予測フラグを計算して、
    　計測時間帯別に、前記修正した稼働／非稼働別に、前記気象データを入力、前記電気使用量データを出力とした回帰分析を行い、計測時間帯別稼働／非稼働別電気使用量予測モデルを作成して、
    　前記計測時間帯別稼働／非稼働別電気使用量予測モデルに、前記気象予報値を入力して、電気使用量予測値と、前記電気使用量予測値の確率分布と、を計算して、
    　前記電気使用量予測値と、前記稼働／非稼働予測フラグと、前記計測時間帯別稼働／非稼働別ベースラインと、前記ベース電気使用量と、を用いて、電気使用量予測値と、空調等、照明等、ベースの用途別電気使用量予測値と、を計算させる請求項１に記載のプログラム。     On the computer,
    Using the weather forecast value obtained from the outside, the weather data, and the electricity usage data, the electricity usage forecast value, the occurrence probability distribution of the electricity usage forecast value, air conditioning, lighting, etc. It is a program that calculates the electricity usage forecast value for each base application,
    The operating / non-operating flag, the operation time / non-operating baseline according to the measurement time zone, and the base electricity usage amount are calculated using claim 1,
    The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operational baseline by measurement time zone, which is a set of measurement dates and times that are non-operational in the baseline by operation / non-operational time period, Set it up and modify the working / non-working flag,
    Enter the day of the week, holiday, time zone, etc. of the measurement date and time, create an operation regression model with the modified operation / non-operation flag as output,
    In the operating regression model, enter the day of the week, holiday and time zone to be predicted, calculate the operating / non-operating prediction flag,
    Regression analysis using the weather data as input and the electricity usage data as output for each modified operation / non-operation for each measurement time period, and creating an electricity usage prediction model for each operation time / operation period do it,
    Input the weather forecast value into the electricity usage amount prediction model for each operation time / non-operation time, calculate the electricity usage prediction value, and the probability distribution of the electricity usage prediction value,
    Using the electricity usage amount predicted value, the operation / non-operation prediction flag, the measurement time zone operation / non-operation baseline, and the base electricity usage amount, the electricity usage prediction value and the air conditioning The program according to claim 1, which calculates a base usage-specific electricity usage predicted value such as lighting, etc.
  3. 　コンピューターに、
    　空調の設定温度の変更による省エネ量を計算させるプログラムであって、
    　事前に空調設定温度の変化量を設定しておき、
    　前記気象データと、前記電気使用量データと、前記計測時間帯別稼働／非稼働別ベースラインと、を用いて、
    　前記計測時間帯別稼働／非稼働別ベースラインが発生した時の気温を冷房と暖房が切り替わる気温閾値として、
    　前記気温閾値より計測日時の気温が高い場合は、前記気温から前記空調設定温度の変化量を引いた値で、引いた値が前記気温閾値より小さくなる場合は、値を前記気温閾値に補正した値を省エネ計算用気温として、
    　前記気温閾値より前記計測日時の気温が低い場合は、前記気温から前記空調設定温度の変化量を加えた値で、加えた値が前記気温閾値より大きくなる場合は、値を前記気温閾値に補正した値を省エネ計算用気温として、
    　請求項１で導出した前記計測時間帯別稼働／非稼働別電気使用量推計モデルに、前記気象データの気温を前記省エネ計算用気温に置換したデータを入力して計算した省エネ計算用電気使用量推計値と、
    　請求項１で導出した前記計測時間帯別稼働／非稼働別電気使用量推計モデルに、前記気象データを入力して計算した電気使用量推計値を用いて、
    　前記電気使用量推計値と、前記省エネ計算用電気使用量推計値と、の差を、空調の設定温度の変更による省エネ量として計算させる請求項１又は請求項２に記載のプログラム。     On the computer,
    A program that calculates the amount of energy saved by changing the temperature setting for air conditioning.
    Set the change amount of air conditioning set temperature in advance,
    Using the meteorological data, the electricity usage data, and the baseline by operation / non-operation by measurement time zone,
    As the temperature threshold at which the cooling and heating are switched, the temperature when the baseline by operation / non-operation is generated by the measurement time zone,
    When the temperature at the measurement date / time is higher than the temperature threshold value, the value obtained by subtracting the amount of change in the air conditioning set temperature from the temperature, and when the subtracted value is smaller than the temperature threshold value, the value is corrected to the temperature threshold value. The value is the temperature for energy saving calculation.
    When the temperature at the measurement date and time is lower than the temperature threshold, a value obtained by adding the amount of change in the air conditioning set temperature to the temperature, and when the added value is greater than the temperature threshold, the value is corrected to the temperature threshold. As the temperature for energy saving calculation,
    Electricity consumption for energy saving calculation calculated by inputting data obtained by replacing the temperature of the meteorological data with the temperature for energy saving calculation in the electricity usage amount estimation model classified by operation time / non-operation time derived in claim 1 Estimated value,
    Using the electricity usage estimation value calculated by inputting the weather data in the electricity usage amount estimation model by operation time / non-operation time derived in claim 1,
    3. The program according to claim 1, wherein a difference between the estimated amount of electricity used and the estimated amount of electricity used for energy saving calculation is calculated as an energy saving amount by changing a set temperature of air conditioning.
  4. 　コンピューターに、
    　合理的な電力デマンドの目標値を計算させるプログラムであって、
    　事前に、異常値の発生確率を定めておき、
    　前記電気使用量データの３０分間隔の合計値と、前記気象データの３０分間隔の平均値と、を請求項１の手法に用いて、計測時間帯別稼働／非稼働別電気使用量推計モデルを計算し、
    　前記電気使用量データの３０分間隔の合計値の最大値を２倍した値を、現状の電力デマンド値として、
    　前記計測時間帯別稼働／非稼働別電気使用量推計モデルから計算される電気使用量推計値の最大値を２倍した値を、達成可能なデマンド値として、
    　前記計測時間帯別稼働／非稼働別電気使用量推計モデルを作成する際に、推計値の確率分布を計算して、前記確率分布から、前記異常値の発生確率が閾値となる推計値を計算し、その値を２倍した値を、達成容易なデマンド値として、
    　それぞれのデマンド値が前記現状の電力デマンド値より大きい場合は、前記現状の電力デマンド値に補正して計算させる請求項１から３のいずれか１項に記載のプログラム。     On the computer,
    A program that calculates a reasonable target value for power demand,
    Predetermining the probability of occurrence of abnormal values in advance,
    The total value of 30 minutes intervals of the electricity usage data and the average value of 30 minutes intervals of the meteorological data are used in the method of claim 1, and the electricity usage amount estimation model by operation time period / non-operation time is used. Calculate
    As a current power demand value, a value obtained by doubling the maximum value of the total value of 30 minutes intervals of the electricity usage data,
    As a demand value that can be achieved, a value obtained by doubling the maximum value of the electricity usage estimated value calculated from the electricity usage estimation model classified by operation / non-operation according to the measurement time period,
    When creating the electricity usage estimation model by operation time period / non-operation time, calculate the probability distribution of the estimated value, and calculate the estimated value from which the occurrence probability of the abnormal value becomes a threshold from the probability distribution Then, double the value as a demand value that is easy to achieve,
    The program according to any one of claims 1 to 3, wherein when each demand value is larger than the current power demand value, the current power demand value is corrected and calculated.
  5. 　コンピューターに、
    　電力デマンドの内訳を表示させるプログラムであって、
    　前記電気使用量データの３０分間隔の合計値と、前記気象データの３０分間隔の平均値と、を請求項１の手法に用いて、空調等、照明等、ベースの用途別電気使用量を計算し、
    　前記電気使用量データの３０分間隔の合計値及び、前記空調等、照明等、ベースの用途別電気使用量をそれぞれ２倍した値を、電力デマンド値及び、空調等、照明等、ベースの用途別電力デマンド値として、
    　前記電力デマンド値を降順にソートして、
    　積み上げ棒グラフを用いて、前記電力デマンド値の用途別内訳を表示させる請求項１から４のいずれか１項に記載のプログラム。     On the computer,
    A program that displays a breakdown of power demand,
    By using the total value of 30 minutes intervals of the electricity usage data and the average value of 30 minutes intervals of the meteorological data in the method of claim 1, the electricity usage by base such as air conditioning, lighting, etc. Calculate
    The total value of the electricity usage data at intervals of 30 minutes and the value of electric power demand value and the base usage such as lighting such as air conditioning etc. As another power demand value,
    Sort the power demand values in descending order,
    The program according to any one of claims 1 to 4, wherein a breakdown by use of the power demand value is displayed using a stacked bar graph.
  6. 　コンピューターに、
    　残業等の所定時間外労働による追加的な電気使用量を計算させるプログラムであって、
    　請求項１を用いて前記稼働／非稼働フラグと、前記計測時間帯別稼働／非稼働別電気使用量推計モデルと、前記計測時間帯別稼働／非稼働別ベースラインと、を計算して、
    　前記計測時間帯別稼働／非稼働別ベースラインの中で非稼働である計測日時の集合である計測時間帯別非稼働ベースラインの最大値より、電気使用量が小さい計測日時は、非稼働にして、稼働／非稼働フラグを修正して、
    　前記計測日時の曜日、祝日、時間帯等を入力、前記修正した稼働／非稼働フラグを出力とした稼働回帰モデルを作成して、
    　前記稼働回帰モデルに、前記計測日時の曜日、祝日、時間帯等を入力して、建物の定期的な稼働／非稼働フラグを計算して、
    　前記定期的な稼働／非稼働フラグが、非稼働である計測日時において、前記修正した稼働／非稼働フラグが稼働である場合に、所定時間外労働としてフラグを立てて、
    　前記計測時間帯別非稼働電気使用量推計モデルを用いて、前記所定時間外労働の計測日時における非稼働電気使用量を推計して、
    　前記所定時間外労働フラグが立っている計測日時における電気使用量と、前記所定時間外労働の計測日時における非稼働電気使用量と、の差を、所定時間外労働による追加的な電気使用量として計算させる請求項１から５のいずれか１項に記載のプログラム。     On the computer,
    A program that calculates additional electricity usage due to overtime work such as overtime,
    Using the operation / non-operation flag according to claim 1, calculating the electricity usage estimation model by operation / non-operation according to measurement time period, and the baseline according to operation / non-operation according to measurement time period,
    The measurement date and time when the amount of electricity used is smaller than the maximum value of the non-operation baseline by measurement time zone, which is a set of measurement dates and times that are non-operational in the baseline by operation time / non-operation time, is set to non-operation. Correct the operating / non-operating flag,
    Enter the day of the week, holiday, time zone, etc. of the measurement date and time, create an operation regression model with the modified operation / non-operation flag as output,
    In the operation regression model, enter the day of the week, holiday, time zone, etc. of the measurement date and time, calculate the regular operation / non-operation flag of the building,
    In the measurement date and time when the periodic operation / non-operation flag is non-operation, when the corrected operation / non-operation flag is operation, a flag is set as overtime for a predetermined time,
    Using the non-operating electricity usage estimation model by measurement time zone, estimating the non-operating electricity usage at the measurement date of the overtime work,
    The difference between the amount of electricity used at the measurement date and time when the overtime work flag is set and the amount of non-operating electricity used at the measurement date and time of the overtime work is defined as an additional amount of electricity used by overtime work. The program according to any one of claims 1 to 5, wherein the program is calculated.
PCT/JP2019/020788 2018-06-09 2019-05-20 Energy-saving diagnostic program WO2019235272A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2019566853A JP6756425B2 (en) 2018-06-09 2019-05-20 Energy saving diagnostic program

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2018121316 2018-06-09
JP2018-121316 2018-06-09

Publications (1)

Publication Number Publication Date
WO2019235272A1 true WO2019235272A1 (en) 2019-12-12

Family

ID=68770339

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2019/020788 WO2019235272A1 (en) 2018-06-09 2019-05-20 Energy-saving diagnostic program

Country Status (2)

Country Link
JP (1) JP6756425B2 (en)
WO (1) WO2019235272A1 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110988556A (en) * 2019-12-20 2020-04-10 上海市建筑科学研究院有限公司 Diagnosis method for architectural lighting socket support in non-operation period
CN112116496A (en) * 2020-09-27 2020-12-22 施耐德电气（中国）有限公司 Configuration method and device of energy consumption diagnosis rules
KR20220157080A (en) * 2021-05-20 2022-11-29 숭실대학교산학협력단 Method and apparatuses for prediction for predicting electrical generation of special day
US20230169426A1 (en) * 2021-11-30 2023-06-01 Ncr Corporation Demand response resource scheduling

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013065087A (en) * 2011-09-15 2013-04-11 Panasonic Corp Power saving evaluation system, power saving evaluation method, server and program
JP2016127622A (en) * 2014-12-26 2016-07-11 株式会社デンソー Power demand prediction system
JP2017050971A (en) * 2015-09-01 2017-03-09 株式会社Ｌｏｏｏｐ Power consumption prediction control apparatus, power consumption prediction method, and power consumption prediction program

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013065087A (en) * 2011-09-15 2013-04-11 Panasonic Corp Power saving evaluation system, power saving evaluation method, server and program
JP2016127622A (en) * 2014-12-26 2016-07-11 株式会社デンソー Power demand prediction system
JP2017050971A (en) * 2015-09-01 2017-03-09 株式会社Ｌｏｏｏｐ Power consumption prediction control apparatus, power consumption prediction method, and power consumption prediction program

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110988556A (en) * 2019-12-20 2020-04-10 上海市建筑科学研究院有限公司 Diagnosis method for architectural lighting socket support in non-operation period
CN112116496A (en) * 2020-09-27 2020-12-22 施耐德电气（中国）有限公司 Configuration method and device of energy consumption diagnosis rules
CN112116496B (en) * 2020-09-27 2023-12-08 施耐德电气（中国）有限公司 Configuration method and device of energy consumption diagnosis rules
KR20220157080A (en) * 2021-05-20 2022-11-29 숭실대학교산학협력단 Method and apparatuses for prediction for predicting electrical generation of special day
KR102536500B1 (en) * 2021-05-20 2023-05-26 숭실대학교 산학협력단 Method and apparatuses for prediction for predicting electrical generation of special day
US20230169426A1 (en) * 2021-11-30 2023-06-01 Ncr Corporation Demand response resource scheduling

Also Published As

Publication number Publication date
JPWO2019235272A1 (en) 2020-06-25
JP6756425B2 (en) 2020-09-16

Similar Documents

Publication Publication Date Title
WO2019235272A1 (en) Energy-saving diagnostic program
CN110097297B (en) Multi-dimensional electricity stealing situation intelligent sensing method, system, equipment and medium
CN108700851B (en) System, method and cloud-based platform for predicting energy consumption
Chakhchoukh et al. Electric load forecasting based on statistical robust methods
Fiot et al. Electricity demand forecasting by multi-task learning
CN110969285B (en) Prediction model training method, prediction device, prediction equipment and medium
Chan et al. A new approach to characterizing and forecasting electricity price volatility
CN109726872B (en) Energy consumption prediction method and device, storage medium and electronic equipment
JP2003242212A (en) Apparatus and method for determining day of the week with similar utility consumption profile
JP2009294969A (en) Demand forecast method and demand forecast device
Park et al. A framework for baseline load estimation in demand response: Data mining approach
Harmer et al. Using calibrated energy models for building commissioning and load prediction
JP7114956B2 (en) Power demand calculation device and program
US20210334914A1 (en) System and method for determining power production in an electrical power grid
CN107730097B (en) Bus load prediction method and device and computing equipment
US20180225779A1 (en) System and method for determining power production in an electrical power grid
JP2017169289A (en) Power prediction system, power prediction method, and program
JP2017034797A (en) Baseline load estimation device and baseline load estimation program
CN116937569A (en) Intelligent energy storage method and device for photovoltaic power generation and electronic equipment
CN114879619A (en) Digital workshop energy optimization method and system
Ramos et al. Using diverse sensors in load forecasting in an office building to support energy management
Chatzos et al. Data-driven time series reconstruction for modern power systems research
KR20160074325A (en) Electricity Demand Index (EDI) Forecasting System with respect to Weather Condition Change
Nikolaev et al. Application of the cybernetic approach to price-dependent demand response for underground mining enterprise electricity consumption
García-Díaz et al. Competitive models for the spanish short-term electricity demand forecasting

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2019566853

Country of ref document: JP

Kind code of ref document: A

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19816065

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19816065

Country of ref document: EP

Kind code of ref document: A1