Computer-implemented method for determining energy load distribution in system

Abstract

A computer-implemented method for determining an energy load distribution in a system includes: receiving total energy load data for the system, where the total energy load data includes individual energy load data for a plurality of assets; determining at least one energy load peak value in the total energy load data; decomposing the total energy load data into individual energy load data of a plurality of assets; the decomposed individual energy load data for the plurality of assets is provided for further processing.

Description

Technical Field

[0001] The present invention relates to a computer-implemented method for determining the energy load distribution in a system, to the use of decomposed individual energy load data of multiple systems in an energy management system, to the use of decomposed individual energy load data of multiple systems on a graphical user interface, to a data processing system, to a computer program product and to a computer-readable medium. Background Art

[0002] The energy consumption of a building or the energy demand of a building plays a key role in efficiency considerations for the operation of a building. A building may be a commercial, residential and / or industrial building. Energy management systems for such buildings are well known in the state of the art and deal with a number of different entities in such buildings.

[0003] It is now clear that there is also a need to provide methods for managing such buildings energetically. Summary of the invention

[0004] In view of the above, an object of the present invention is to provide a method for determining energy load distribution in a system, and in particular an object of the present invention is to provide an improved method for determining energy load distribution in a system.

[0005] These and other objects, which become apparent on reading the following description, are solved by the subject-matter of the independent claims. The dependent claims mention preferred embodiments of the invention.

[0006] In one aspect of the present disclosure, a computer-implemented method for determining energy load distribution in a system is provided, comprising the following steps: receiving total energy load data for the system, wherein the total energy load data includes individual energy load data for a plurality of assets; determining at least one energy load peak in the total energy load data; decomposing the total energy load data into individual energy load data for a plurality of assets; providing the decomposed individual energy load data for the plurality of assets for further processing. The decomposed individual energy load data may be advantageously used to control one or more of the plurality of assets in order to eliminate energy load peaks. This may be advantageous in analyzing the system. This may be advantageous in providing further insights into the system. This may advantageously allow control strategies to be developed for a plurality of assets.

[0007] In an embodiment of the method, determining at least one energy load peak may be based on a pattern recognition algorithm. This may be advantageous in terms of efficient detection of energy load peaks. This may advantageously prevent the system from being ahead of further energy load peaks or prevent a detected energy load peak from continuing. The information of energy peaks may further be used to select new hardware in the system or to change the production plan. The information of energy peaks may generally give new insights. An energy manager or energy management system may use these insights to develop specific steps (e.g., change the production plan).

[0008] In an embodiment of the method, determining at least one energy load peak may be based on a trained machine learning algorithm. This may be advantageous in terms of efficient detection of energy load peaks. This may advantageously prevent the system from tripping before further energy load peaks or prevent a detected energy load peak from continuing.

[0009] In an embodiment of the method, decomposing the total energy load data may include an estimation algorithm. This may be advantageous in the absence of individual energy load data for one or more assets.

[0010] In an embodiment of the method, decomposing the total energy load data may include receiving one or more individual energy loads of a plurality of assets via a measuring device. This is advantageous in terms of the accuracy and quality of the determined decomposed energy load data.

[0011] In an embodiment of the method, the method may further include providing a tracking record of the plurality of assets, wherein the tracking record is based on the decomposed individual energy load data of the plurality of assets. This may be advantageous in terms of user-friendliness and analytical capabilities.

[0012] In an embodiment of the method, the method may further include marking a plurality of assets as controllable assets and uncontrollable assets. This may be advantageous because it further provides the possibility of controlling controllable assets. This may be advantageous in terms of efficiency, responsiveness and automation capabilities.

[0013] In an embodiment of the method, the plurality of assets may include one or more of the following: heater, EV charger, battery, HVAC (heating, ventilation, and air conditioning equipment).

[0014] Another aspect of the present disclosure relates to the use of decomposed individual energy load data of multiple systems being implemented in an energy management system through the method described above.

[0015] Another aspect of the present disclosure relates to the use of decomposed individual energy load data for multiple systems being implemented on a graphical user interface through the method described above.

[0016] Another aspect of the present disclosure relates to a data processing device comprising means for performing the method described above.

[0017] Another aspect of the present disclosure relates to a computer program product comprising instructions which, when executed by a computer, cause the computer to perform the method described above.

[0018] A final aspect of the present disclosure relates to a computer-readable medium comprising instructions which, when executed by a computer, cause the computer to perform the method described above.

[0019] definition

[0020] The term energy load profile must be understood broadly and relates to one or more individual energy loads of one or more assets over time. Preferably, the energy load profile includes individual energy loads of multiple assets. The energy load profile may be described in a table or a graph. The energy load profile may include data values ​​relating to energy load consumption and corresponding time. The time may be seconds, hours, days, weeks, months and / or years. The term energy load must be understood broadly and relates to energy load consumption or power of an asset. The energy load is preferably expressed in watts. Energy or power may come from one or more energy sources. Energy may include electrical energy, thermal energy and / or mechanical energy.

[0021] The term system must be understood broadly and refers to any entity comprising one or more assets, wherein the assets consume electrical energy. Preferably, the system may be a residence, a factory, an office building. The system may include subsystems, such as plant 1 and plant 2.

[0022] The term asset must be understood broadly and refers to any electrical energy consuming entity related to the system. The asset may include lights, heaters, HVAC, EV chargers in a building. The term asset is not limited to the examples given.

[0023] The term total energy load data must be understood broadly, in particular, essentially relates to the entire energy load consumption of the system. For example, the total energy load data is obtained by an integral meter of a building, wherein the meter may be provided by an electricity supplier via an interface of the integral meter. The total energy load data consists of a plurality of energy consumers (i.e. assets). The total energy load data comprises at least an energy load value and a time value, preferably a curve of energy values ​​varying with time. The total energy load data may be represented in the form of numbers, tables, graphs, etc. The total energy load data may be received from an energy management system of each system via an interface.

[0024] The term individual energy load data is to be understood broadly and preferably relates to the individual energy load consumption of an asset.

[0025] The term peak energy load is to be understood broadly and preferably relates to the highest power consumed within a period of time.

[0026] The term decomposed individual energy load data of multiple assets must be understood broadly and relates to one or more individual energy load consumption of one or more corresponding assets. In other words, the decomposed individual energy load data relates to the distribution of the total energy load data of the individual assets, for example, at the energy load peak at 11:13 am on Tuesday, the heater consumed 30% of the energy load, the light consumed 20% of the energy load and the EV charger consumed 50% of the energy load.

[0027] The term pattern recognition algorithm must be understood broadly and may relate to any pattern analysis algorithm configured to determine energy load peaks in total energy load data. Pattern recognition algorithms may include classification algorithms, artificial intelligence algorithms, machine learning algorithms. The term pattern recognition algorithm preferably relates to techniques that automatically recognize certain patterns and regularities in total energy load data.

[0028] The term machine learning algorithm must be understood broadly and may preferably relate to a trained machine learning model. The machine learning model may be trained based on labeled data, for example, using a historical energy demand log for a building. The machine learning algorithm may preferably include a decision tree, a naive Bayesian classification, a nearest neighbor, a neural network, a convolutional neural network, a generative adversarial network, a support vector machine, a linear regression, a logistic regression, a random forest, and / or a gradient boosting algorithm.

[0029] The term estimation algorithm is understood broadly and may relate to an algorithm configured to estimate individual energy load and total energy load data (individual energy load of one asset or individual energy load of multiple assets) based on information of one or more assets in the system and / or some measurements of one or more assets. The estimation algorithm may be a decomposition algorithm. Such a decomposition algorithm may work according to the following steps: First, all individual assets (e.g., a building: refrigerators, washing machines, monitors, etc.) may be identified within the system, and then, the energy consumption of all assets as well as the entire building may be collected frequently (e.g., every second) over a longer period of time (e.g., 1 year). These data may be used to train the AI ​​model (in the training phase). The algorithm may then extract relevant so-called features or characteristics from the collected data. Examples for these features may be power usage patterns, spectrum, voltage fluctuations, time of day, etc.

[0030] With historical data and extracted features, ML models can be trained to learn to classify and identify assets responsible for specific energy patterns. Finally, in the decomposition process phase, the current total energy demand can be taken, specific features and characteristics studied, and the ML model can decompose and predict the load distribution into individual assets.

[0031] The term measuring device is understood broadly and relates to any device configured for measuring the energy load consumption of an asset. The measuring device may be an electric meter, a shunt resistor or similar device. The measuring device may include an interface for transmitting the measured values ​​to another entity, such as a central processing unit.

[0032] The term trace record is understood broadly and refers to the individual load data broken down over time for multiple assets. The trace record can be displayed on a screen for further analysis. The trace record can include numbers, tables, energy load values.

[0033] The term controllable asset must be understood broadly and relates to assets that can control their own energy load consumption. Controllable assets may include EV chargers. A user may decide to start charging an electric vehicle via an EV charger late at night based on the individual energy load data broken down from multiple assets. For some assets, the power load can be dynamically controlled, at least to some extent, by an external system, such as an energy management system. EV chargers can be turned on and off, and some EV chargers even allow continuous control, i.e. the charging power can take any value within a minimum and maximum charging rate range. HVAC systems can also provide some flexibility. For example, the power supplied to a heating or cooling system can be dynamically adjusted based on the temperature of the room. In contrast, some assets cannot be controlled from the outside, such as traditional electric heating systems that can only be controlled manually.

[0034] The term non-controllable assets must be understood broadly and relates to assets that are generally not able to control their own energy load consumption. Non-controllable assets may include safety equipment, such as fire alarm equipment that must be permanently activated. Marking assets as controllable and non-controllable assets may be beneficial to improve the efficiency of the method. This information may allow the user or a superior control to switch assets that are dependent on the decomposed individual load data.

[0035] The term energy management system is understood broadly and refers to any software used to analyze, control or predict system energy behavior.The energy management system may provide one or more of the following data: total energy load data, one or more measured individual energy loads of one or more assets.

[0036] Units and / or devices according to one or more example embodiments may be implemented using hardware, software, and / or a combination thereof. For example, hardware devices may be implemented using processing circuits including, but not limited to, processors, central processing units (CPUs), controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field programmable gate arrays (FPGAs), systems on chips (SoCs), programmable logic units, microprocessors, or any other device capable of responding to and executing instructions in a defined manner.

[0037] The unit or device may include one or more interface circuits. In some examples, the interface circuit may include a wired or wireless interface connected to a local area network (LAN), the Internet, a wide area network (WAN), or a combination thereof. The functions of any given device or unit of the present disclosure may be distributed between multiple units or devices connected via an interface circuit.

[0038] Any disclosure and embodiment described herein relates to the above listed methods, systems, apparatus, computer program elements, and vice versa. Advantageously, the benefits provided by any embodiment and example are equally applicable to all other embodiments and examples, and vice versa.

[0039] As used herein, "determine" also includes "initiate or cause determination", "generate" also includes "initiate or cause generation", and "provide" also includes "initiate or cause determination, generation, selection, sending or receiving". "Initiate or cause execution of an operation" includes any processing signal that triggers a computing device to execute a corresponding operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Hereinafter, the present disclosure is exemplarily described with reference to the accompanying drawings, in which

[0041] Figure 1 A flow chart illustrating an example method for determining energy load distribution in a system;

[0042] Figure 2 The use of an example method for determining an energy load profile in conjunction with an energy management system is shown;

[0043] Figure 3 An example data processing apparatus comprising means for performing the method described above is shown;

[0044] Figure 4 An exemplary diagram showing the total energy load of a system;

[0045] Figure 5 An exemplary graph showing decomposed individual energy load data for a plurality of assets;

[0046] Figure 6 An exemplary graph showing tracking records for multiple assets is shown. DETAILED DESCRIPTION

[0047] Figure 1 A flow chart of an example method for determining energy load distribution in a system is shown.

[0048] Step S100 includes receiving total energy load data for a system, wherein the total energy load data includes individual energy load data for a plurality of assets. The total energy load data may be received from an energy management system. The total energy load data may be received from an interface of an energy metering device, wherein the energy metering device measures the total energy load of the system. The system may be an office building. Alternatively, the system may be a factory building, a residence, or a similar building. In the example below, the total energy load data may include a refrigerator, an EV charger, and a heater as a plurality of assets.

[0049] Step S110 includes determining at least one energy load peak value in the total energy load data. Determining at least one energy load peak value can be based on a pattern recognition algorithm. In the following example, the pattern recognition algorithm is based on a machine learning algorithm. The machine learning algorithm can be trained using labeled historical log data of the total energy data of the system.

[0050] Once the energy load peak is determined, step S120 can be initiated. Alternatively, step S120 can be performed continuously. Step S120 includes decomposing the total energy load data into individual energy load data for multiple assets. The decomposition can be based on an estimation algorithm. The decomposition can include receiving one or more individual measured loads from one or more assets. It should be noted that the method includes a combination of estimation of individual energy loads and measurement of individual energy loads. In this example, the individual load of the refrigerator is determined by an electrical measuring device disposed on the refrigerator, wherein the electrical measuring device also includes an interface configured to transmit the individual load data to a data processing device that executes the method described herein. The individual energy loads of other assets (i.e., heaters and EV chargers) are determined by an estimation algorithm based on the total energy data.

[0051] Step S130 includes providing the decomposed individual energy load data of multiple assets for further processing. The decomposed individual energy load data can be displayed on a screen for further analysis. The decomposed individual energy load data can be used to control one or more controllable assets. In this example, in order to read the peak energy load used, the EV charger can be turned off. The decomposed individual energy load data can serve the continuous control of the behavior of one or more assets, or serve the future planning of the control of one or more assets. It should be noted that the method described is automatically operated and the decomposed individual energy load data can also be used for multiple assets of the automatic system, for example in an automatically operated energy management system.

[0052] The method may also include providing a tracking record of the plurality of assets, wherein the tracking record is based on the decomposed individual energy load data of the plurality of assets. The method may also include marking the plurality of assets as controllable assets and controllable assets. This may be advantageous because it increases the efficiency of the energy management system.

[0053] Figure 2 The use of an example method for determining an energy load profile in conjunction with the energy management system 100 is shown.

[0054] The energy management system 100 continuously provides system total energy load data. The total energy load data is continuously searched for energy load peaks by the determination unit 110. Once a peak is detected by the determination unit 110, the decomposition unit 120 decomposes the total energy load data into individual energy load data for a plurality of assets. The decomposed individual energy load data for a plurality of assets are then provided for further processing. Further processing may include deriving control data 130 from the decomposed individual energy load data. Further processing may also include transmitting the decomposed individual energy load data to a statistics unit 140, which is configured to analyze the decomposed individual energy data in order to provide further insights to a user 150.

[0055] Figure 3 An example data processing device 200 including means for performing the above method is shown. The means may include a central processing unit, a desktop pc, a workstation, a cloud application, a smartphone. The means may include one or more interfaces for data transmission.

[0056] Figure 4 An exemplary graph 300 of the total energy load of a system is shown. The horizontal axis 310 shows time and the vertical axis 320 shows the total energy load of all assets. The graph 300 includes two energy load peaks 330 and 340.

[0057] Figure 5 An exemplary graph 400 of decomposed individual energy load data for a plurality of assets is shown. The two peaks 410 and 420 of the total energy load of the exemplary graph 400 are decomposed into individual energy loads 430, 440, 450, and 460.

[0058] Figure 6 An exemplary graph 500 is shown of a tracking record of a plurality of assets 510 and 520. As can be seen from the figure, the assets include a freezer, a charger, and a heater, each of which is associated with their respective individual loads.

[0059] Below, a summary of the main effects and main advantages of the present invention is provided:

[0060] The maximum energy load peak, i.e. the highest power consumed in a given period, can be a major contributor to the energy bill. However, due to the large number of individual energy loads and their dynamic nature, it is often impossible to understand which assets contribute to how much the energy load peaks. The solution described in the present invention is a method to automatically identify energy load peaks in the total energy load data and automatically determine the decomposition of the load contribution of each individual asset to the energy load peak. These insights can enable the user to understand the cause of the peak and enable the user to take specific actions. These individual energy loads can be used to control individual assets in order to reduce the energy load peaks.

[0061] The total energy load data or total energy load demand profile for commercial, residential and industrial buildings consists of many different individual energy loads. These individual energy loads may be associated with various assets, such as HVAC systems, EV chargers or similar equipment. Typically, the energy load demand curve is not flat, but may have a dynamic behavior such that it varies over time and occasionally contains total energy load peaks.

[0062] The total energy cost depends not only on the total energy consumed, but also on the maximum peak, i.e. the highest power consumed in a given period, which is an important factor. Therefore, to understand which assets contribute to the peak and how much, a high rate of interest is needed. However, due to the large number of individual energy loads and their dynamic nature over time, this is usually impossible through manual inspection. In other words, an automated computer-implemented method is needed to determine the individual energy loads of multiple assets in order to flatten the actual energy load curve and avoid future peaks.

[0063] The proposed method automatically identifies energy demand peaks and automatically determines a breakdown of the load contribution of each individual asset to the peak. Furthermore, the method can highlight flexible loads, ie loads for which (at least some) control is feasible.

[0064] Insights from the breakdown of peak loads to the individual assets themselves are beneficial to control systems, e.g. buildings and the assets themselves. These insights enable users to further understand the causes of peaks and therefore put users in a position to take specific actions. For example, if it is observed that most peaks are caused by intensive use of EV chargers, users can schedule charging processes to be more spread out throughout the day and / or consider purchasing a battery energy storage system. Furthermore, this insight can be automatically processed to generate control commands for assets.

[0065] The idea involves methods to automatically identify peaks in total energy demand and automatically determine the load demand that breaks down the total load demand into individual assets. The method can be built on top of existing energy management and / or energy monitoring systems. The method can include one or more of the following steps:

[0066] set up:

[0067] Step 1: Mark assets whose loads are (at least to some extent) controllable as “controllable”.

[0068] Step 2: Connect the energy decomposition module to the energy management / monitoring system.

[0069] Step 3: Connect the pattern recognition system to the energy management / monitoring system.

[0070] Repeat:

[0071] Step 4: Pattern Recognition The system continuously searches for a "peak pattern" in the total load.

[0072] Step 5: Whenever a peak pattern is found, the energy decomposition module decomposes the total load into individual energy loads and allocates these loads to physical assets (electric heaters, air conditioners, EV chargers...).

[0073] Step 6: The system keeps track of all assets and their contributions to the different peaks over time. In addition to this record, two things are calculated:

[0074] a. Continuously average statistics for each physical asset.

[0075] b. The proportion of the total load contributed by the "controllable" load.

[0076] The individual steps are explained in more detail below. The numerical values ​​and asset examples are for illustrative purposes only.

[0077] Step 1 (Mark assets as “controllable”):

[0078] For some assets, electrical loads can be controlled dynamically, at least to some extent, by external systems:

[0079] EV chargers can be turned on and off, and some EV chargers even allow continuous control, where the charging power can take any value within a minimum and maximum charging rate range. HVAC systems also offer some flexibility. For example, the power supplied to the heating or cooling system can be dynamically adjusted based on the temperature of the room.

[0080] Assets that fall into this category are marked as “controllable.” Conversely, some assets cannot be controlled from the outside, such as traditional electric heating systems that can only be controlled manually.

[0081] Step 2 (Energy decomposition):

[0082] Energy disaggregation is a technique that takes as input a total energy demand signal and disaggregates it into individual energy signals for each asset. If the energy management / monitoring system already sub-meters some of the building's electrical loads, the energy disaggregation can be combined with the existing sub-metering data to obtain a complete and more accurate sub-metering. Finally, if all individual energy loads are monitored, the method of the present invention does not require energy disaggregation.

[0083] Step 3 & 4 (Pattern Recognition):

[0084] Pattern recognition is a technique for automatically identifying certain patterns and regularities in data. In the method of the present invention, pattern recognition is used to identify peak patterns in energy demand data. Pattern recognition systems are usually based on machine learning, and they can be trained on labeled data, for example, historical energy demand logs of a building.

[0085] Step 5 (Energy decomposition):

[0086] Each time a peak pattern is identified in step 4, the total energy load is decomposed into individual energy loads using the energy decomposition module introduced in step 2. Energy decomposition is applied to the cumulative energy profile over time for which the pattern recognition system has identified the peak. The combination of peak pattern detection and energy decomposition provides insights such as that charging of electric vehicles is the main contributor to a day's peak. If a user or energy management system observes this more frequently, the user or energy management system is now aware of the cause of the peak and can thus initiate specific preventive measures (e.g., deriving control commands for the assets).

[0087] Step 6 (Save tracking records):

[0088] Finally, a trace of all detected peaks (step 4) and the respective decomposed energy profiles of each asset (step 5) are saved in a database. This trace can be used to provide explanations and statistics over a longer period of time to users (e.g., energy managers).

[0089] The method can be implemented on top of existing energy management / monitoring systems.

[0090] With this capability, the energy management system will be able to automatically provide users (i.e. energy managers) with insights into questions such as: Can a single or small number of assets be identified as the main contributor to a peak? Or do all assets contribute almost equally to the peak? How often and with what regularity do peaks occur? How much of the peak is controllable? Is each peak caused by a similar group of assets?

[0091] Reference numerals

[0092] S100 receives the total energy load data of the system

[0093] S110 Determine at least one energy load peak

[0094] S120 decomposition of total energy load data

[0095] S130 provides decomposed total energy load data

[0096] 100 Energy Management System

[0097] 110 Determine the unit

[0098] 120 decomposition units

[0099] 130 Control Data

[0100] 140 analysis units

[0101] 150 users

[0102] 200 data processing equipment

[0103] 300 Total Energy Load Diagram

[0104] 310 horizontal axis

[0105] 320 vertical axis

[0106] 330, 340 Energy load peak

[0107] 400 Total Energy Load Diagram

[0108] 410, 420 Energy load peak

[0109] 430, 440, 450, 460 individual energy loads

[0110] 500 Tracking Record Chart

[0111] 510, 520 assets

Claims

1. A computer-implemented method for determining energy load distribution in a system, comprising: receiving total energy load data for the system, wherein the total energy load data includes individual energy load data for a plurality of assets; determining at least one energy load peak value in the total energy load data; decomposing the total energy load data into the individual energy load data of the plurality of assets; The decomposed individual energy load data for the plurality of assets is provided for further processing. 2 . The method according to claim 1 , wherein determining at least one energy load peak is based on a pattern recognition algorithm.

3. The method according to claim 1 or 2, wherein determining at least one energy load peak is based on a trained machine learning algorithm.

4. A method according to any one of the preceding claims, wherein decomposing the total energy load data comprises an estimation algorithm.

5. The method of any one of the preceding claims, wherein decomposing the total energy load data comprises receiving one or more individual energy loads of the plurality of assets by means of a measuring device.

6. The method of any one of the preceding claims, further comprising providing a tracking record of the plurality of assets, wherein the tracking record is based on the decomposed individual energy load data of the plurality of assets.

7. The method of any preceding claim, further comprising marking the plurality of assets as controllable assets and non-controllable assets.

8. The method of any of the preceding claims, wherein the plurality of assets comprises one or more of the following: a heater, an EV charger, a battery, a HVAC. 9 . Use of decomposed individual energy load data of a plurality of assets received by the method according to claim 1 in an energy management system.

10. Use of decomposed individual energy load data of a plurality of assets received by the method according to any one of claims 1 to 8 on a graphical user interface.

11. A data processing device comprising means for executing the method according to any one of claims 1 to 8.

12. A computer program product comprising instructions which, when the program is executed by a computer, cause the computer to perform the method according to any one of claims 1 to 8.

13. A computer-readable medium comprising instructions, which, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 8.

Images