Energy program communication control system and method based on load shape analysis

A predictive model using load shape analysis and demographic data classifies energy customers to target energy efficiency program communications, improving program participation rates and reducing resource waste by focusing on likely participants.

JP7590868B2Active Publication Date: 2024-11-27ORACLE INT CORP
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
JP2020542004
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2018-02-01
Filing Date
2019-01-29
Publication Date
2024-11-27
Estimated Expiration
2039-01-29

AI Technical Summary

Technical Problem

Utility companies face challenges in effectively targeting energy efficiency programs to customers based on their likelihood to participate, leading to inefficient communication and resource waste, as conventional systems lack the ability to distinguish between likely and unlikely participants.

Method used

A predictive model using load shape analysis and demographic data is employed to classify energy customers, allowing targeted communication of energy efficiency program information to those most likely to participate, thereby reducing unnecessary communication and resource waste.

Benefits of technology

The model effectively identifies likely participants, optimizing communication efforts and reducing costs and customer irritation by focusing on a targeted subset of customers, thus enhancing the effectiveness of energy efficiency program promotions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007590868000001
    Figure 0007590868000001
  • Figure 0007590868000002
    Figure 0007590868000002
  • Figure 0007590868000003
    Figure 0007590868000003
Patent Text Reader

Abstract

Systems, methods, and other embodiments associated with predicting participation in an energy efficiency program are described. A method embodiment includes obtaining load data for an energy customer and determining an empirical load shape for the energy customer based on the obtained load data. A prescribed load shape that most closely matches the empirical load shape is selected for the energy customer. A data structure is generated to include the prescribed load shape, demographic data, and / or land parcel data. By applying a trained predictive model to the data structure, it is determined that the energy customer is more likely to participate in the energy efficiency program than a different energy customer. Communication of information regarding the energy efficiency program is controlled by assigning a higher priority to a communication to the energy customer than a second communication to the different energy customer.
Need to check novelty before this filing date? Find Prior Art

Description

[Background technology]

[0001] background Utility companies and other entities offer a variety of energy efficiency programs aimed at motivating customers to consume energy efficiently. Each such program has different measures offered as incentives to energy customers. For example, financial incentives or other savings measures may be offered to purchasers of efficient appliances under energy efficient product programs. More specifically, electric utility companies may offer rebates directly to consumers of energy efficient products to offset a portion of the premium typically charged by retailers for energy efficient products. Electric utility companies may offer a discount rate to be charged on electricity delivered to residences occupied by energy customers who purchase such energy efficient products. Government tax credits are also available to customers who purchase energy efficient products.

[0002] With so many programs available and such a variety of different approaches offered under these programs, utility companies lack a practical way to inform energy customers about available programs. Furthermore, utility companies have traditionally been unable to distinguish those energy customers who are likely to participate in an energy efficiency program from those who are not. Computer systems may allow data records for energy customers to be sorted or grouped into classes based purely on generalized data identifying the type of structure identified in the data records for the energy customers. For example, energy customers may be grouped into a subgroup of customers who reside in single-family homes and a subgroup of customers who reside in multi-family apartment buildings. However, the propensity of customers within each such subgroup to participate in energy efficiency programs may vary widely. As a result, conventional computer systems have been unable to filter data records within a general classification for energy customers who are likely to participate in energy efficiency programs from data records of customers who are not likely to participate in order to control communications about such programs. Instead, information about all available programs has traditionally been transmitted to all energy customers via electronic or physical communication channels. Communicating information about every program to every energy customer wastes computational and network resources for information that is communicated electronically (e.g., email, push notifications, etc.). Communicating information about every program to every energy customer wastes natural and financial resources for information that is communicated physically (e.g., via the postal service, etc.). And, regardless of the mode of communication, bombarding energy customers with information about irrelevant energy efficiency programs can irritate those energy customers. Annoyed energy customers may ignore all such information received in the future, even if that information pertains to a program for which they qualify. Summary of the Invention [Means for solving the problem]

[0003] overview In one embodiment, a non-transitory computer readable medium is disclosed that stores computer executable instructions that, when executed by at least one processor of a computing system, cause the computing system to perform the following operations: obtaining load data for an energy customer from a database, the load data indicative of power consumed by the energy customer at different times, the operations further including generating an empirical load shape for the energy customer based on the obtained load data, the empirical load shape representing power consumed by the energy customer over a period of time, the period of time including the different times in the obtained load data, the operations further including selecting, from among a plurality of established load shapes stored in a storage device, a prescribed load shape for the energy customer that most closely matches the empirical load shape; and generating, for the energy customer, a combination of the prescribed load shape, demographic data, and parcel data, or the prescribed load shape and demographic data and parcel data. and determining that the energy customer is more likely to participate in the energy efficiency program than a different energy customer by applying a predictive model to the data structure, the predictive model being trained at least in part based on the defined load shapes for the different energy customers, the operations further including controlling communication of information regarding the energy efficiency program, the controlling communication of the information including controlling communication of the information by prioritizing a communication of the information to the energy customer over a second communication of the information to the different energy customer as a result of determining that the energy customer is more likely to participate in the energy efficiency program than the different energy customer.

[0004] In another embodiment of the non-transitory computer-readable medium, the predictive model is a binary classifier, and the computer-executable instructions, when executed by the at least one processor of a computing system, cause the at least one processor to use the binary classifier to classify the energy customers into classes of customers likely to participate in the energy efficiency program and classify the different energy customers into classes of customers unlikely to participate in an energy efficiency program.

[0005] In another embodiment of the non-transitory computer readable medium, the binary classifier is a logistic regression model that classifies the energy customer from the different energy customer.

[0006] In another embodiment of the non-transitory computer-readable medium, the load data obtained from the database includes a total electricity consumption by the energy customer for each of a plurality of different days, and the empirical load shape represents a pattern of total electricity consumption per day by the energy customer.

[0007] In another embodiment of the non-transitory computer readable medium, the predictive model is an empirically derived algorithm derived from a combination of one or more of the following data: load data, demographic data, and parcel data for a training set of customers, the customers in the training set of customers including (i) a plurality of participating customers who participated in the energy efficiency program and (ii) a plurality of non-participating customers who did not participate in the energy efficiency program.

[0008] In another embodiment of the non-transitory computer readable medium, a predictive model is trained based on a dataset including a plurality of established load shapes for the plurality of different energy customers.

[0009] In another embodiment of the non-transitory computer-readable medium, computer-executable instructions are executed to cause a computing system to prevent a second communication of information relating to the energy efficiency program to the different energy customer.

[0010] In another embodiment, a computing system is disclosed.The computing system comprises at least one processor coupled to at least one memory; and a data management module stored on a non-transitory computer readable medium and including instructions that, when executed by the at least one processor, cause the computing system to retrieve load data for energy customers from a database, the load data indicative of power consumed by the energy customer at different times; the computing system further comprises an analysis module stored on a non-transitory computer readable medium and including instructions that, when executed by the at least one processor, cause the computing system to perform the following operations: generating an empirical load shape for the energy customer based on the retrieved load data, the empirical load shape representing power consumed by the energy customer over a period of time; and selecting a defined load shape for the energy customer from among a plurality of established load shapes stored in a storage device that most closely matches the empirical load shape for the energy customer. selecting a load shape; generating a data structure for the energy customer including a defined load shape, demographic data, parcel data, or a combination of the defined load shape and at least one of the demographic data and parcel data; and determining that the energy customer is more likely to participate in an energy efficiency program than a different energy customer by applying a predictive model to the data structure, the computing system further comprising a transmission control module stored on a non-transitory computer readable medium and including instructions that, when executed by the at least one processor, cause the computing system to control transmission of information regarding the energy efficiency program, wherein controlling the transmission of the information includes controlling the transmission of the information by prioritizing a transmission of the information to the energy customer over a second transmission of the information to the different energy customer as a result of determining that the energy customer is more likely to participate in the energy efficiency program than the different energy customer.

[0011] In another embodiment of the computing system, the predictive model is a binary classifier and the analysis module further includes instructions that, when executed by the at least one processor, cause the computing system to classify the energy customers into a class of customers who are likely to participate in the energy efficiency program and classify the different energy customers into a class of customers who are unlikely to participate in an energy efficiency program.

[0012] In another embodiment of the computing system, the binary classifier is a logistic regression model that classifies the energy customer from the different energy customers.

[0013] In another embodiment of the computing system, the database is in communication over a communications network with a meter provided to measure load data obtained for a residence associated with the energy customer, and stores measurements received over the communications network and obtained by the meter.

[0014] In another embodiment of the computing system, the load data stored by the database and retrieved by the data management module includes total electricity consumption of the energy customer for different days, and instructions of the analysis module generate an empirical load shape to represent a pattern of total electricity consumption per day by the energy customer.

[0015] In another embodiment of the computing system, the predictive model is an empirically derived algorithm derived from a combination of one or more of the following data: load data, demographic data, and parcel data for a training set of customers, the customers in the training set including (i) a plurality of participating customers who participated in the energy efficiency program and (ii) a plurality of non-participating customers who did not participate in the energy efficiency program.

[0016] In another embodiment of the computing system, a predictive model is trained based on a dataset that includes a plurality of established load shapes for the plurality of different energy customers.

[0017] In another embodiment of the computing system, the transmission control module further includes instructions that, when executed by the at least one processor, cause the computing system to prevent a second transmission of information relating to the energy efficiency program to the different energy customer.

[0018] In another embodiment, a computer-implemented method is disclosed that includes obtaining load data for an energy customer from a database, the load data indicative of power consumed by the energy customer at different times, the method further includes generating an empirical load shape for the energy customer based on the obtained load data, the empirical load shape representing power consumed by the energy customer over a period of time, the period of time including the different times in the obtained load data, the method further includes selecting, from among a plurality of established load shapes stored in a storage device, a prescribed load shape that most closely matches the empirical load shape for the energy customer, and generating an empirical load shape for the energy customer based on the prescribed load shape, the demographic data, generating a data structure including parcel data or a combination of a defined load shape and at least one of demographic data and parcel data; determining that the energy customer is more likely to participate in an energy efficiency program than a different energy customer by applying a predictive model to the data structure; and controlling the dissemination of information regarding the energy efficiency program, wherein controlling the dissemination of information includes controlling the dissemination of information by prioritizing a dissemination of the information to the energy customer over a second dissemination of the information to the different energy customer as a result of determining that the energy customer is more likely to participate in the energy efficiency program than the different energy customer.

[0019] In another embodiment of the computer-implemented method, the predictive model is a binary classifier, and determining that the energy customer is more likely to participate in the energy efficiency program than the different energy customer includes classifying the energy customer into a class of customers who are more likely to participate in the energy efficiency program and classifying the different energy customer into a class of customers who are less likely to participate in the energy efficiency program.

[0020] In another embodiment of the computer-implemented method, a predictive model is trained based on a dataset including a plurality of established load shapes for the plurality of different energy customers.

[0021] In another embodiment of the computer-implemented method, the load data obtained from the database includes a total electricity consumption by the energy customer for each of a plurality of different days, and determining an empirical load shape for the energy customer includes determining a pattern of total electricity consumption per day by the energy customer.

[0022] Another embodiment of the computer-implemented method further includes preventing a second communication of information regarding the energy efficiency program to the different energy customer.

[0023] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate various systems, methods, and other embodiments of the present disclosure. It will be understood that the boundaries of elements illustrated in the figures (e.g., boxes, groups of boxes, or other shapes) represent certain embodiments of the boundaries. In some embodiments, an element may be realized as multiple elements, or the multiple elements may be realized as one element. In some embodiments, an element shown as an internal component of another element may be realized as an external component, and vice versa. Additionally, elements may not be drawn to scale. [Brief description of the drawings]

[0024] [Figure 1] FIG. 1 illustrates a schematic diagram of an embodiment of an energy distribution network including a computing system for predicting energy customer participation in an energy efficiency program. [Diagram 2]FIG. 1 illustrates a flow diagram that generally illustrates an embodiment of a method for predicting the propensity of an energy customer to participate in an energy efficiency program using a computing system. [Diagram 3] FIG. 1 illustrates an example computing device configured and / or programmed according to one or more of the example systems and methods described herein. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] Detailed Description Described herein is a computerized system and method that manipulates data records for energy customers to identify load shapes that represent the measured energy consumption patterns in the data records. Each of the identified load shapes is matched to an established load shape that most accurately represents the energy consumption pattern in the respective data record. A predictive model linked to the established load shape that most accurately represents the energy consumption pattern is applied to the information in the data record. Application of the predictive model results in the designation of the data record as being associated with an energy customer that is likely to participate in an energy efficiency program or associated with an energy customer that is unlikely to participate. These data records may be filtered to limit communication of information regarding the energy efficiency program to the energy customer using at least a threshold propensity to participate.

[0026] Until now, utility companies have relied on manufacturers and retailers of energy-saving products to offer these products to their energy customers, using energy efficiency as a selling point. Otherwise, utility companies have instituted their own information campaigns using traditional approaches such as sending out bulk physical literature to all energy customers. However, such approaches have demonstrated limited effectiveness.

[0027] In an effort to address at least some of the above shortcomings, the present disclosure involves computer-implemented techniques and algorithms, including supervised machine learning, for "training" a predictive model. The trained predictive model attempts to distinguish between data records associated with energy customers who have a propensity to participate in an energy efficiency program and data records associated with energy customers who are unlikely to participate. The predictive model is trained using a training set of data that includes prior participation data, available demographic and parcel data, and / or collected data, such as energy consumption data measured using smart meters or other meters, for each energy customer in the training set. The training set includes data records for both energy customers known to have participated and energy customers known not to have participated in the particular energy efficiency program for which the predictive model is trained. The relative proportions of data records for energy customers known to have participated and energy customers known not to have participated may be approximately the same as the proportions in the underlying data set. Having been trained in this manner, the predictive model may be applied to data records for energy customers not in the training set who have an unknown propensity to participate in the energy efficiency program. Therefore, the trained predictive model can predict whether other non-participating energy customers are likely to participate in such programs in the future.

[0028] The training involves developing a model that distinguishes between data records for energy customers who are likely to participate in an energy efficiency program and data records for energy customers who are unlikely to participate in the program based at least in part on one or more of empirically determined load shapes of the dwelling associated with the energy customer, information about the dwelling, such as square footage, type of heat (e.g., forced air, hot water, heat pump, etc.), and condition of the dwelling, demographic information of the energy customer, and previous participation history of the energy customer. An embodiment involves developing a model that distinguishes between data records for energy customers who are likely to participate and data records for energy customers who are unlikely to participate based at least in part on a combination of (i) empirically determined load shapes of the dwelling associated with the energy customer, and (ii) information about the dwelling, such as square footage, type of heat (e.g., forced air, hot water, heat pump, etc.), and condition of the dwelling. The empirically determined load shapes may be categorized into a load shape that most closely matches a plurality of prescribed load shapes and a predictive model that is trained using the prescribed load shapes. Predictive models can be trained separately for energy efficiency programs that are specific to one or more of residence type, demographic information, and prior participation history by using corresponding training sets of data.

[0029] Once trained, the predictive model can be applied to rank and / or filter the data records of non-participants by their predicted propensity to participate in the energy efficiency program in the future. Non-participants are energy customers who are eligible to participate in the energy efficiency program but are not currently participating. To apply the predictive model, the daily energy usage patterns in the data records for the energy customers are characterized as one of a small number, but more than one, of established load shapes. The characterization involves finding the established load shape that most closely matches the empirical load shape. The established load shapes can be defined in a database and can include a generic representation of the N most commonly occurring load shapes, where N can be any positive integer value greater than 1 (e.g., N can be equal to 2, 3, 4, 5, 6, 7, 8, etc.). The trained predictive model is then applied to the data for the non-participants using the established load shape that most closely matches the empirical load shape as a variable of the predictive model. Thus, the predictive model uses the most closely matching established load shape as one variable. One or more additional variables, such as housing information, heat type, housing condition, etc., are also input into the predictive model to distinguish non-participants who are likely to participate from non-participants who are unlikely to participate.

[0030] The use of energy customers' daily load shapes to train predictive models has shown significant predictive capabilities beyond those of other features. Utility companies can conserve resources by targeting information about energy efficiency programs to a smaller subset (e.g., less than all) of energy customers who are predicted to be more likely to participate than customers outside of the subset. Additionally, utility companies can reduce marketing costs and costs to customer satisfaction that come from receiving irrelevant or unwelcome marketing.

[0031] 1, an exemplary embodiment of a system 100 associated with analyzing data records to predict energy customer participation in an energy efficiency program is shown. System 100 is shown and described herein for use with an electric power distribution system 105 for purposes of clearly illustrating the present technology. However, system 100 may be utilized in connection with any other utility or consumable supplier, such as, for example, a natural gas distribution system.

[0032] The electrical power distribution system 105 includes an establishment 110, which may be a generating station, substation, distribution station, or any other supplier of electricity operated by or on behalf of an electric utility company. A network including transmission lines 115 supported by one or more towers 120 conducts electricity between the establishment 110 and residences 125 associated with energy customers. The network may include other transmission equipment such as transformers and switches, although such equipment is omitted from FIG. 1 for clarity. Additionally, while the residences 125 are shown as single-family residential homes, other examples of residences 125 include any single-family residential home or residential apartment complex, condominium, apartment, etc., commercial office buildings, retail stores, industrial plants, government buildings, or any other structures associated with energy customers and supplied with electrical energy by an electric utility company.

[0033] The residence 125 is equipped with a meter 130 that measures the amount of electricity consumed by the residence 125 over time. An embodiment of the meter 130 may be a so-called "smart meter." The smart meter optionally includes a network communication module (not shown) that communicates data over a communication network 135 at regular intervals, such as monthly, quarterly, etc. The communication network 135 may include a wide area network ("WAN") such as the Internet, a local area network ("LAN") including residential or local networking equipment such as wireless routers, switches and modems, or a combination of a WAN and a LAN.

[0034] The transmitted data may be data used to accurately bill the energy customer associated with the residence 125 for the actual amount of electricity consumed at the residence 125. The transmitted data may also optionally be stored in a database 140 available to the system 100. The data transmitted by the smart meter may include, but is not limited to, a portion of the load data 145 indicative of the amount of electricity consumed by the energy customer at the residence 125 at different times, and optionally all such data. For example, the data transmitted by the smart meter may include data that unambiguously identifies, or may be used to determine, the total electricity consumed during a week, day, hour, etc.; the length of time for which electricity consumption is reported, such as a month, quarter, etc.; kilowatt-hours; and the like.

[0035] For embodiments of meters 130 that may lack the ability to communicate data over communications network 135, data that would be communicated by the smart meter over communications network 135 can be manually read and recorded in database 140. Whether the data is communicated over communications network 135 or manually recorded in database 140, the data can be associated with data records corresponding to different energy customers.

[0036] In one embodiment, system 100 is a computing system that includes an application or a collection of distributed applications for an enterprise organization. The application and computing system 100 may be configured to operate with or be implemented as a cloud-based networking system, a software as a service (SaaS) architecture, or other type of networked computing solution. In one embodiment, system 100 is a centralized server-side application that provides at least the functionality disclosed herein and is utilized by many users via computing devices / terminals that communicate with system 100 (acting as a server) through computer / communication network 135.

[0037] In some embodiments, one or more of the components described herein are configured as program modules stored on at least one non-transitory computer-readable medium, the program modules being configured with stored instructions that, when executed by at least one processor, cause a computing device to perform the corresponding function(s) described herein.

[0038] The system 100 of FIG. 1 includes a database 140 that stores at least a portion of the data used by the system 100 described herein. The database 140 can reside on a non-transitory computer readable medium or can include data stored by an array of non-transitory computer readable media, which may be comprised of multiple storage devices. According to yet other embodiments, the information contained in the database 145 described with respect to the illustrated embodiment may be obtained from multiple different databases. Furthermore, any portion of the data contained in the database 140 and utilized by the system 100 may optionally be obtained from a remote database that is external to and does not form part of the system 100. Such an external database may be available to the system 100 via the communications network 135 or a local data communications link. For simplicity, the database 140 shown in FIG. 1 is described herein as including data that is utilized as a factor by the system 100 to predict the propensity of energy customers to participate in energy efficiency programs. Such predictions are accomplished by applying a trained predictive model to data records associated with the energy customers.

[0039] Load data 145 is among the data included in the data records in database 140. Load data 145 is indicative of electricity consumed by the energy customer at residence 125 at different times. Load data 145 can be utilized by system 100 to determine the total electricity consumption by the energy customer at residence 125, for example, in kilowatt-hours per day, for each of a number of different days within a period of time (e.g., a month, a three-month period, etc.).

[0040] Database 140 also includes customer data 150 specific to residence 125 and / or specific to that energy customer associated with a data record. For example, in one embodiment, customer data 150 may include one or more, or a combination of, the following types of data that may be part of a data record corresponding to residence 125: Demographic data available to the electric utility company. For example, demographic data may include the number of adults residing in the residence 125 associated with the energy customer, the number of children residing in the residence 125 associated with the energy customer, whether the energy customer owns or rents the residence 125, etc. Parcel data or other data describing characteristics of the residence 125. For example, the parcel data may include the square footage, number of floors, or other indicators of the size of the residence 125 associated with the energy customer. The type of heat used to heat the residence 125 during cold seasons (e.g., electric, gas, fuel oil, wood, etc.), the type of residence (e.g., single-family home, apartment complex, etc.), may also be included in the parcel data. Miscellaneous data that may be useful in predicting an energy customer's propensity to participate in an energy efficiency program may also be included in customer data 150. For example, the number of other programs and / or initiatives (excluding the program of interest) in which the energy customer is already a participant may be stored in database 140 as customer data 150. The energy customer's past participation (or lack thereof) in energy efficiency programs and / or initiatives may be a useful indicator of whether the energy customer is generally receptive to such programs and / or initiatives.

[0041] A library 155 containing a number of established load shapes is also included in the database 140. The electrical load is the total power used by the dwelling 125 for a given time (e.g., in real time, for a particular time, for a day, etc.). The electrical load changes over time in response to changes in the operation of electrical equipment such as lighting, heating, ventilation, air conditioning, computer and office equipment, furnaces, and other industrial equipment. The electrical load may be represented by measured data that defines a graphical curve, other line graph, plot, or other graphical representation as a function of time. The curve represents the load shape and graphically reveals the pattern of power consumption.

[0042] Each established load shape in library 155 is a data structure that defines a generalized representation of the load shape expected (e.g., based on previously observed data) to be exhibited by one or more particular types of dwellings 125. Being generalized representations, the established load shapes may not accurately reflect empirical load shapes developed based on actual electricity consumption data, such as load data 145. Instead, the generalized shapes may be shapes that represent an idealized load shape having a smoothed pattern that represents characteristics unique to the electricity consumption patterns of different buildings.

[0043] For example, a regular daily repeating load shape 160 with equally spaced rectangular peaks as shown in FIG. 1 may indicate an office building that is closed on weekends. Power consumption is generally uniform at a high level during the workday (e.g., Monday-Friday), hence the rectangular peaks. On weekends, when the office building is closed, power consumption is generally uniform at a low level, forming valleys between the rectangular peaks. Thus, the load shape 160 may be associated with residences 125, such as office buildings and schools, that are closed on weekends but are in use during normal hours during the week.

[0044] Conversely, an established load shape for a residential home occupied by two adults who work normal work hours during the day (e.g., 8:00AM-4:30PM) may be the inverse of load shape 160. During work days (Monday-Friday), while the home is vacant, the power consumed generally levels out at a low level, resulting in valleys for the work days. On weekends, when the resident adults are home longer during the day, the power consumption generally levels out at a high level, forming rectangular peaks between valleys for the weekend days. Different established load shapes can be included in library 155 to represent other common load shapes.

[0045] Although an embodiment of the load shape 160 is described as a weekly load shape, the disclosure is not so limited. Another embodiment of the load shape 160 is a daily load shape. Instead of a curve of power consumed versus time over the course of a week (e.g., sampled at a daily frequency), a daily load shape is a curve of power consumed versus time over the course of a 24-hour day. Therefore, the sampling frequency of a daily load curve can be much higher than one sample recorded each day. For example, a daily load shape can be shown as power consumed on a scale of seconds, minutes, or hours. Additionally, the daily load shape can be compared to established load shapes that have patterns specific to work days, have patterns specific to weekend days, and are based on overall usage according to the day for which the empirical load shape is generated. For example, an empirical load shape for work days can be compared to an established load shape specific to work days. Similarly, a load shape for power consumed during weekend days can be compared to an established load shape specific to weekend days.

[0046] To ensure availability of a sufficient sample size of data used to train the predictive model, the number of established load shapes in the library 155 may be limited. For example, the number may be limited to a number less than the number of all known load shapes. In an embodiment, the library 155 may be limited to include N of the most commonly occurring load shapes, where N may be any positive integer value greater than 1 (e.g., N may be equal to 2, 3, 4, 5, 6, 7, 8, etc.). Each established load shape may be associated with a type of day (e.g., workday, weekend, etc.), a category of week (e.g., season, holiday, etc.). In other words, the empirical load shape may be compared with different established load shapes specific to different time periods, with the different established load shapes appropriately input into the predictive model.

[0047] System 100 also includes a data management module 165. Data management module 165 serves as an interface with database 140. At a minimum, load data 145 corresponding to the energy customer may be retrieved from database 140 by data management module 165. However, other information, such as customer data 150 and / or established load shapes in library 155, may also be retrieved by the data management module as described herein.

[0048] The analysis module 170 uses data acquired by the data management module 165, including load data 145 for the energy customer, to determine and generate an empirical load shape 175 for the energy customer based on the acquired load data 145. The empirical load shape 175 represents a pattern of actual electricity consumption by the energy customer over a period of time as measured by the meter 130. As an example, the empirical load shape 175 may represent daily electricity consumption at the residence 125 associated with the energy customer over the course of a month. The analysis module 170 then identifies an established load shape from the set of load shapes stored in the library 155, such as load shape 160 in FIG. 1, that most closely matches the empirical load shape 175. The established load shape 160 that is the most closely matched is selected by the analysis module 170 as what is hereinafter referred to as the “defined load shape 160.” The defined load shape 160 is used by the system 100 to determine an appropriate predictive model 180 to be applied by the system 100 to a data structure corresponding to the energy customer. The results of applying the predictive model to the energy customer's data structure determine the energy customer's propensity to participate in energy efficiency programs.

[0049] The predictive model may be trained using a training set of data for energy customers that are classified into each of a variety of different established load shapes. As an example, the entire population of energy customers included in the training set of data is classified into a number of subsets. Each subset may include energy customers whose daily or other periodic load shapes are classified into a different one of a number of established load shapes. The load shape for each of the energy customers in the training set of data may be categorized into a subset that corresponds to a closest matching load shape among the N different established load shapes. A data structure, such as a feature vector for each energy customer may be created to include the closest matching established load shape along with other information associated with the energy customer, such as information about the residence, heat type, and residence conditions.

[0050] Each feature vector is assigned a binary label indicating whether the respective energy customer is a participant in the program or policy of interest. The program or policy of interest is a program or policy in which the energy customer is predicted to participate. A logistic regression analysis is performed using the feature vectors for each of the subsets to develop a logistic regression model or other suitable algorithm as a predictive model. For example, the trained predictive model may be a binary classifier that classifies the energy customers into one of two classes: (i) a class of customers likely to participate in the energy efficiency program, or (ii) a class of customers unlikely to participate in the energy efficiency program. Regardless of the type of predictive model, any scalar values, error constants, formula structures, etc. of the predictive model should distinguish feature vectors having different binary labels with a reasonable degree of accuracy. Thus, the predictive model may be trained to be a function of the established load shape and at least one other input variable associated with the energy customer, such as information about the residence, demographic data, heat type, and residential conditions, etc.

[0051] After the initial training of the predictive model is completed, subsequent refinements can occur. As predictions are made for energy customers using the trained predictive model, the scalar values ​​or error constants, or even the structure, of the predictive model can be modified from the results of the initial training. Predictions resulting from use of the predictive model, and data indicating whether the energy customer ultimately chose to participate in the energy efficiency program, can be used to update the predictive model.

[0052] The analytics module 170 generates a data structure for the energy customer to predict the propensity of the energy customer to participate in an energy efficiency program. The generated data structure may be a vector including at least demographic data and / or land parcel data corresponding to the residence 125 corresponding to or associated with the energy customer. A trained predictive algorithm 180 obtained from the database 140 is applied to the data structure to determine the propensity of the energy customer to participate in an energy efficiency program. The predictive results may be embodied in a data structure output by the analytics module 170.

[0053] A transmission control module 185 of the system 100 receives the prediction results and controls transmission of information regarding the energy efficiency program based on the prediction results. Controlling the transmission of the information may involve updating or otherwise modifying the database 140 to tag data records associated with the energy customer with a trend designation. For example, if the prediction results indicate that the energy customer is more likely to participate in an energy efficiency program than a different energy customer, the transmission control module 185 may designate the energy customer as a recipient of the information. The updated record with this designation may subsequently be identified by searching the data record for the tag and may be utilized by a system separate from the system 100. The identified data record with the tag is therefore identified as a recipient to which an electronic message containing information regarding the energy efficiency program should be transmitted.

[0054] The electronic transmission of the electronic message may be addressed to an email address, a number where texts can be received, or other electronic contact information for the energy customer. The electronic transmission may then be transmitted over communications network 35 based on the designation of the energy customer as a recipient. According to other embodiments, a separate system may utilize the updated records and may process letters, brochures, bills, or other physical correspondence containing information about the energy efficiency program based on the designation in the updated records. According to other embodiments, a separate system may utilize the updated records and control what content is presented to the energy customer, for example, when the energy customer logs into a user account with a utility company to pay a bill.

[0055] Regardless of the manner of communication, communication of such information to the energy customer may be prioritized over communication of such information to a different energy customer who may have a predicted propensity to participate in the energy efficiency program that is lower than a threshold or at least lower than the predicted propensity of the energy customer. For such embodiments, communication control module 185 may output, modify, or otherwise update data structures that prevent communication of information regarding the energy efficiency program to the different energy customer.

[0056] A flow diagram illustrating a method 200 for predicting an energy customer's propensity to participate in an energy efficiency program using system 100 is illustrated in Figure 2. In response to system 100 receiving a request to predict the propensity of the energy customer, data management module 165, in step 205, retrieves load data 145 for the residence 125 corresponding to the energy customer using database 140. The load data may be used to identify or determine an empirical load shape, for example, per day.

[0057] Using the load data 145 obtained by the data management module 165, the analysis module 170 determines an empirical load shape for a defined time period (e.g., daily). The empirical load shape represents the power consumed in the residence 125 over that time period. As described above, the time period can be a day, a week, a two-week period, a month, or another desired time period. The analysis module 170 compares the determined empirical load shape to the established load shapes in the library 155 in step 215. The analysis module 170 selects the established load shape that most closely matches the empirical load shape in step 220. For example, data values ​​from the empirical load shape are compared to data values ​​from one of the established load shapes (and repeated for the other load shapes), with the matching function being based at least on a threshold of the difference in the values ​​based on any distance metric, such as Euclidean distance. The selected established load shape is identified as the defined load shape in step 220.

[0058] A data structure, such as a feature vector for the energy customer, is generated in step 225. The generated data structure includes values ​​for parameters, such as an established load shape that most closely matches an empirical load shape associated with the energy customer, demographic data, and / or parcel data, that form the input to a predictive model. Using the analytics module 170 in step 230, the trained predictive model is applied to the data structure to determine that the energy customer is more likely to participate in an energy efficiency program than a different energy customer. Determining the relative propensity of the energy customer to the different energy customer may involve comparing respective propensity values ​​for the energy customer and the different energy customer. According to an alternative embodiment, the determined propensity of the energy customer may be compared to a threshold value that separates likely participants from unlikely participants. According to yet other embodiments, the propensity may simply be determined for each of the energy customer and the different energy customer. These propensity values ​​may then be presented to a user and manually compared by the user to determine the relative propensity of the energy customer to the different energy customer.

[0059] Based on the output from analysis module 170, transmission control module 185 transmits, modifies, or generates data to control transmission of information regarding the energy efficiency program to the energy customer in step 235. For example, transmission control module 185 can prioritize transmission of the information to the energy customer over transmission of the information to the different energy customer. According to an alternative embodiment, transmission control module 185 can cause information regarding the energy efficiency program to be transmitted to the energy customer rather than to the different energy customer.

[0060] 3 illustrates an example of a computing device 300 configured and / or programmed with one or more of the example systems and methods described herein and / or equivalents. An illustrative example of a computing device 300 may be a computer 315 including a processor 320, a memory 335, and an I / O port 345 operably connected by a bus 325. In an embodiment, the computer 315 includes logic circuitry of a data management module 165, an analysis module 170, and a transmission control module 185 configured to facilitate the system 100 and / or methods described with respect to FIGS. 1 and 2. In different embodiments, the logic circuitry of the data management module 165, the analysis module 170, and the transmission control module 185 may be implemented in hardware, a non-transitory computer-readable medium 305 having instructions stored thereon, firmware, and / or combinations thereof. While the logic of the data management module 165, the analysis module 170, and the transmission control module 185 are illustrated as hardware components attached to the bus 325, it should be understood that in other embodiments the logic of one or more of these modules may be implemented in the processor 320, stored in the memory 335, or stored on the disk 355.

[0061] In an embodiment, the logic of the data management module 165, the analysis module 170 and the transmission control module 185, or the computer 315, is a means (e.g., structure: hardware, non-transitory computer readable medium, firmware) for performing the described operations. In some embodiments, the computing device 300 may be a server operating in a cloud computing system, a server configured in a Software as a Service (SaaS) architecture, a smartphone, a laptop, a tablet computing device, and the like.

[0062] The means may be implemented, for example, as an application specific integrated circuit (ASIC) that is programmed to implement rule-based source sequencing for allocation. The means may also be implemented as stored computer-executable instructions that are temporarily stored in memory 335 and subsequently executed by processor 320 and provided to computer 315 as data 310.

[0063] The logic of the data management module 165, the analysis module 170, and the transmission control module 185 may also provide the means (e.g., hardware, non-transitory computer-readable medium 305 storing executable instructions, and firmware) for performing rule-based source sequencing for allocation.

[0064] Generally describing an exemplary configuration of computer 315, processor 320 may be a variety of different processors including dual microprocessors and other multi-processor architectures. Memory 335 may include volatile and / or non-volatile memory. Non-volatile memory may include, for example, read-only memory (ROM), programmable read-only memory (PROM), etc. Volatile memory may include, for example, random access memory (RAM), static random-access memory (SRAM), dynamic random access memory (DRAM), etc.

[0065] The disk 355 may be operatively connected to the computer 315, for example, via the I / O interface 340 (e.g., a card, a device) and the I / O port 345. The disk 355 may be, for example, a magnetic disk drive, a solid-state disk drive, a floppy disk drive, a tape drive, a Zip drive, a flash memory card, a memory stick, and the like. Additionally, the disk 355 may be a CD-ROM drive, a CD-R drive, a CD-RW drive, a DVD ROM, and the like. The memory 335 may store processes, such as in the non-transitory computer-readable medium 305, and / or data 310. The disk 355 and / or the memory 335 may store an operating system that controls and allocates resources of the computer 315.

[0066] Computer 315 may interact with input / output (I / O) devices through I / O interface 340 and I / O ports 345. The I / O devices may be, for example, a keyboard, a microphone, a pointing and selection device, a camera, a video card, a display, a disk 355, and network devices 350. The I / O ports 345 may include, for example, serial ports, parallel ports, and USB ports. I / O controller 330 may connect I / O interface 340 to bus 325.

[0067] The computer 315 can operate in a network environment and thus can be connected to a network device 350 via the I / O interface 340 and / or the I / O port 345. Through the network device 350, the computer 315 can interact with a network. Through the network, the computer 315 can be logically connected to a remote computer (e.g., the computer 315 can exist in a distributed computing environment to which a client can connect). The networks with which the computer 315 can interact include, but are not limited to, a local area network (LAN), a new area network (WAN), and other networks.

[0068] In another embodiment, the described methods and / or their equivalents may be implemented using computer-executable instructions. Thus, in one embodiment, a non-transitory computer-readable / storage medium is configured with stored computer-executable instructions that, when executed by a machine(s), cause the machine(s) (and / or associated components) to perform the method. Exemplary machines include, but are not limited to, processors, computers, servers operating in a cloud computing system, servers configured in a Software as a Service (SaaS) architecture, smartphones, and the like. In one embodiment, a computing device is implemented with one or more executable algorithms configured to perform any of the disclosed methods.

[0069] In one or more embodiments, the disclosed methods or their equivalents are performed by either computer hardware configured to perform the methods, or by computer instructions embodied in modules stored on a non-transitory computer-readable medium, the instructions configured as an executable algorithm that, when executed by at least one processor of a computing device, is configured to perform the methods.

[0070] For purposes of simplicity of explanation, the method illustrated in the figures is shown and described as a series of algorithmic blocks, but it should be understood that the method is not limited by the order of the blocks. Some blocks may occur in a different order than shown and described, and / or concurrently with other blocks shown and described. In addition, fewer than all illustrated blocks may be used to implement the example method. Blocks may be combined or separated into multiple operations / components. Furthermore, additional and / or alternative methods may use additional operations not illustrated in the blocks.

[0071] The following includes definitions of selected terms used herein. The definitions include various examples and / or forms of components that fall within the scope of the terms and may be used for implementations. The examples are not intended to be limiting. Both singular and plural forms of the terms may be within the definitions.

[0072] References to "one embodiment," "an embodiment," "one example," "an example," and the like indicate that the embodiment(s) or example(s) so described may include a particular feature, structure, characteristic, property, element, or limitation, but not every embodiment or example necessarily includes that particular feature, structure, characteristic, property, element, or limitation. Moreover, repeated use of the phrase "in one embodiment" does not necessarily refer to the same embodiment, but may.

[0073] A "data structure" or "data object," as used herein, is an organization of data in a computing system that is stored in a memory, storage device, or other computerized system. A data structure may be, for example, any one of a data field, a data file, a data array, a data record, a database, a data table, a graph, a tree, a linked list, and the like. A data structure may be formed from and include many other data structures (e.g., a database includes many data records). Other examples of data structures are possible as well, according to other embodiments.

[0074] "Computer readable medium" or "computer storage medium" as used herein refers to a non-transitory medium that stores instructions and / or data that, when executed, are configured to perform one or more of the disclosed functions. Data may function as instructions in some embodiments. Computer readable media may take forms including, but not limited to, non-volatile and volatile media. Non-volatile media may include, for example, optical and magnetic disks, and the like. Volatile media may include, for example, semiconductor memory, dynamic memory, and the like. Common forms of computer readable media may include, but are not limited to, floppy disks, flexible disks, hard disks, magnetic tapes, other magnetic media, application specific integrated circuits (ASICs), programmable logic devices, compact disks (CDs), other optical media, random access memory (RAM), read only memory (ROM), memory chips or cards, memory sticks, solid state storage devices (SSDs), flash drives, and other media that a computer, processor, or other electronic device can use to function. Each type of media may include stored instructions of an algorithm that, if selected for implementation in an embodiment, is configured to perform one or more of the functions disclosed and / or claimed.

[0075] "Logic" as used herein refers to components implemented using computer or electrical hardware, non-transitory media having stored instructions of executable applications or program modules, and / or combinations thereof to perform any of the functions or operations disclosed herein and / or to cause functions or operations from another logic circuit, method, and / or system to be performed as disclosed herein. Equivalent logic circuits can include firmware, a microprocessor programmed with an algorithm, a discrete logic circuit (e.g., ASIC), at least one circuit, an analog circuit, a digital circuit, a programmed logic device, and a memory device containing algorithmic instructions, any of which can be configured to perform one or more of the disclosed functions. In an embodiment, a logic circuit can include one or more gates, a combination of gates, or other circuit components configured to perform one or more of the disclosed functions. Where multiple logic circuits are described, it may be possible to incorporate multiple logic circuits into one logic circuit. Similarly, where a single logic circuit is described, it may be possible to distribute the single logic circuit among multiple logic circuits. In an embodiment, one or more of these logic circuits are corresponding structures associated with performing the functions disclosed and / or claimed. The choice of which type of logic circuit to implement may be based on system requirements or specifications. For example, if higher speed is a consideration, then hardware would be selected to implement the functions. If lower cost is a consideration, then stored instructions / executable applications would be selected to implement the functions.

[0076] An "operable connection," or a connection in which entities are "operably connected," is a connection in which signals, physical communications, and / or logical communications may be transmitted and / or received. An operable connection may include physical interfaces, electrical interfaces, and / or data interfaces. An operable connection may include different combinations of interfaces and / or connections sufficient to enable operable control. For example, two entities may be operably connected to communicate signals directly to each other or through one or more intermediate entities (e.g., a processor, an operating system, logic circuits, non-transitory computer-readable media). Logical and / or physical communication channels may be used to create an operable connection.

[0077] "User," as used herein, includes, but is not limited to, one or more persons that consume energy from a utility, or a computer or other device operated by or on behalf of such persons, or any combination thereof.

[0078] Although the disclosed embodiments have been illustrated and described in considerable detail, it is not the intention to limit the scope of the appended claims to such detail, or in any way restrictive. Of course, for purposes of describing various aspects of the subject matter, it is not possible to describe every conceivable combination of components or methods. As such, the present disclosure is not limited to the specific details or illustrative examples shown and described. Therefore, the present disclosure is intended to embrace alternatives, modifications, and variations that fall within the scope of the appended claims.

[0079] To the extent that the terms "includes" or "including" are used in the detailed description or the claims, the term "including" is intended to be inclusive in the same manner as the term "comprising" is interpreted when the term "comprising" is used as a transitional word in the claims.

[0080] To the extent the term "or" is used in the detailed description or claims (e.g., A or B), the term "or" is intended to mean "A or B, or both." When applicant intends to indicate "only A or B, but not both," then the phrase "only A or B, but not both" is used. Thus, use of the term "or" herein is inclusive and not exclusive.

Claims

1. A computer readable program storing computer executable instructions which, when executed by at least one processor of a computing system, cause the at least one processor to perform the following operations: obtaining load data for energy customers from a database, the load data indicative of electricity consumed by the energy customers at different times; The operation further comprises: generating an empirical load shape for the energy customer based on the captured load data, the empirical load shape representing a load curve of power consumed by the energy customer versus time over a period of time, the period of time including the different times in the captured load data; The operation further comprises: selecting a prescribed load shape from among a plurality of established load shapes stored in a memory device that most closely matches the empirical load shape for the energy customer; generating a first data structure including a combination of the defined load shape, a binary label indicating whether the energy customer is a participant in an energy efficiency program, demographic data associated with the energy customer, and land parcel data associated with the energy customer; the demographic data includes information indicating whether the energy customer owns a residence, and the parcel data includes an indication of a size of the residence; The operation further comprises: obtaining a load shape for each of a plurality of different energy customers from the database; generating an empirical load shape for each of the plurality of distinct energy customers based on the load shape for each of the plurality of distinct energy customers obtained; generating, for each of the plurality of distinct energy customers, a second data structure including a combination of an empirical load shape for each of the plurality of distinct energy customers, demographic data associated with each of the plurality of distinct energy customers, and land parcel data associated with each of the plurality of distinct energy customers; A predictive model is applied with the defined load shape as a variable of the predictive model, and the operation further comprises: determining a propensity for each of the plurality of different energy customers to participate in the energy efficiency program by applying the second data structure to the predictive model; the predictive model is trained based at least in part on the defined load shape and whether the energy customer is a participant in the energy efficiency program; and The operation further comprises: generating a third data structure for a first customer of the plurality of different energy customers based on a determination that the first customer is likely to participate in the energy efficiency program, the third data structure including an electronic message having information regarding the energy efficiency program; The operation further comprises: transmitting the third data structure to the first customer.

2. The predictive model is a binary classifier including a logistic regression model, and the computer-executable instructions, when executed by the at least one processor of the computing system, cause the at least one processor to: categorizing the plurality of different energy customers into classes of customers likely to participate in the energy efficiency program; or The computer readable program of claim 1 , further comprising: categorizing the plurality of different energy customers into classes of customers unlikely to participate in the energy efficiency program.

3. 3. The computer readable program of claim 1 or 2, wherein the load data obtained from the database includes total electricity consumption by the energy customer and each of the plurality of different energy customers for each of a plurality of different days, and the empirical load shape represents a pattern of total electricity consumption per day by the energy customer and each of the plurality of different energy customers.

4. The computer readable program of any one of claims 1 to 3, wherein the predictive model is trained based on a dataset comprising a plurality of established load shapes for the energy customers.

5. 3. The computer readable program of claim 2, further comprising instructions that, when executed by the at least one processor, cause the at least one processor to prevent communication of the information regarding the energy efficiency program to those customers of the plurality of different energy customers that are unlikely to participate in the energy efficiency program.

6. 1. A computing system, comprising: at least one processor coupled to at least one memory; and a data management module stored on a computer readable medium and including instructions that, when executed by the at least one processor, cause the computing system to retrieve load data for energy customers from a database. the load data indicative of power consumed by the energy customers at different times; The computing system further comprises: an analysis module stored on the computer readable medium and including instructions that, when executed by the at least one processor, cause the computing system to perform the following operations: The operation includes: generating an empirical load shape for the energy customer based on the acquired load data, the empirical load shape representing a load curve of power consumed by the energy customer versus time over a period of time; The operation further comprises: selecting a prescribed load shape from among a plurality of established load shapes stored in a memory device that most closely matches the empirical load shape for the energy customer; generating a first data structure including a combination of the defined load shape, a binary label indicating whether the energy customer is a participant in an energy efficiency program, demographic data associated with the energy customer, and land parcel data associated with the energy customer; the demographic data includes information indicating whether the energy customer owns a residence, and the parcel data includes an indication of a size of the residence; The operation further comprises: obtaining a load shape for each of a plurality of different energy customers from the database; generating an empirical load shape for each of the plurality of distinct energy customers based on the load shape for each of the plurality of distinct energy customers obtained; generating, for each of the plurality of distinct energy customers, a second data structure including a combination of an empirical load shape for each of the plurality of distinct energy customers, demographic data associated with each of the plurality of distinct energy customers, and land parcel data associated with each of the plurality of distinct energy customers; A predictive model is applied with the defined load shape as a variable of the predictive model, and the operation further comprises: determining a propensity for each of the plurality of different energy customers to participate in the energy efficiency program by applying the second data structure to the predictive model; the predictive model is trained based at least in part on the defined load shape and whether the energy customer is a participant in the energy efficiency program; and The operation further comprises: generating a third data structure for a first customer of the plurality of different energy customers based on a determination that the first customer is likely to participate in the energy efficiency program, the third data structure including an electronic message having information about the energy efficiency program; The operation further comprises: transmitting the third data structure to the first customer.

7. The database is associated with the energy customers through a communications network.

7. The computing system of claim 6, in communication with a meter provided to measure the obtained load data for a selected dwelling, and storing measurements received over the communications network and obtained by the meter.

8. 8. The computing system of claim 6 or 7, wherein the predictive model is an empirically derived algorithm derived from one or more combinations of load data, demographic data, and land parcel data for a training set of customers, the customers in the training set of customers including (i) a plurality of participating customers who participated in the energy efficiency program and (ii) a plurality of non-participating customers who did not participate in the energy efficiency program.

9. 1. A computer-implemented method, the method comprising: obtaining load data for energy customers from a database, the load data indicative of electricity consumed by the energy customers at different times; The method further comprises: generating an empirical load shape for the energy customer based on the captured load data, the empirical load shape representing a load curve of power consumed by the energy customer versus time over a period of time, the period of time including the different times in the captured load data; The method further comprises: selecting a prescribed load shape from among a plurality of established load shapes stored in a memory device that most closely matches the empirical load shape for the energy customer; generating a first data structure including a combination of the defined load shape, a binary label indicating whether the energy customer is a participant in an energy efficiency program, demographic data associated with the energy customer, and land parcel data associated with the energy customer; the demographic data includes information indicating whether the energy customer owns a residence, and the parcel data includes an indication of a size of the residence; The method further comprises: obtaining a load shape for each of a plurality of different energy customers from the database; generating an empirical load shape for each of the plurality of distinct energy customers based on the load shape for each of the plurality of distinct energy customers obtained; generating, for each of the plurality of distinct energy customers, a second data structure including a combination of an empirical load shape for each of the plurality of distinct energy customers, demographic data associated with each of the plurality of distinct energy customers, and land parcel data associated with each of the plurality of distinct energy customers; A predictive model is applied with the defined load shape as a variable of the predictive model, and the operation further comprises: determining a propensity for each of the plurality of different energy customers to participate in the energy efficiency program by applying the second data structure to the predictive model; the predictive model is trained based at least in part on the defined load shape and whether the energy customer is a participant in the energy efficiency program; and The method further comprises: generating a third data structure for a first customer of the plurality of different energy customers based on a determination that the first customer is likely to participate in the energy efficiency program, the third data structure including an electronic message having information regarding the energy efficiency program; The method further comprises: transmitting the third data structure to the first customer.

Citation Information

Patent Citations

  • Eco score analytics system

    US20120095794A1

  • Determining likelihood of an individual consumer enrolling in a behavior-based energy efficiency program

    US20150081384A1