Determining process-related energy consumption information for plurality of automated processes

Abstract

The invention makes it possible to determine process-related energy consumption information for a plurality of automated processes in a low-overhead manner based on measurement data acquired for all processes. In this case, the energy consumption change progresses of the plurality of processes are matched with the process information of the respective processes, and the energy consumption changes of the plurality of processes are associated with the respective processes on the basis of the process information of the processes. Energy consumption information for each process is then determined based on a change in energy consumption of the plurality of processes associated with the respective process.

Description

Technical Field

[0001] The present invention relates to a method, an apparatus, a computer program, and a computer program product for determining process-related energy consumption information of a plurality of automated processes. Background Art

[0002] The sustainability of energy places high demands on energy transparency, which must be broken down to the level of each process performed by a device. Understanding how a process consumes energy and how much energy it consumes is crucial for determining the potential to improve energy efficiency and reduce carbon dioxide emissions. Industry is one of the most energy-intensive application areas, which places high demands on energy conservation and energy efficiency.

[0003] One possibility for determining consumption at the level of a single process is to directly perform measurements, that is, to set at least one dedicated sensor (usually a current sensor, supplemented by a voltage sensor if necessary) for each process, and the dedicated sensor collects relevant data for the process. This method has the disadvantage of high overhead, not only in terms of the installation and maintenance costs of the sensors, but also in terms of their energy consumption. In English, this design of directly performing measurements is also referred to as "intrusive load monitoring".

[0004] In addition, there is a possibility of installing only one measuring device for collecting the total current and voltage data of a central measuring point connected to all devices. With the help of an algorithm based on the characteristic behavior of the device, the fingerprint of the device is identified in the energy-related data, and the total energy consumption is broken down to the device level. Modern solutions are based on the structure of the device model, and the device model is trained using a database that contains typical energy-related data of each device. This requires collecting a large amount of data for each device type through direct measurement. The applicability of these solutions in industrial applications is very limited because there is a wide variety of devices in industry, especially in manufacturing processes. Therefore, these solutions based on the typical behavior of the device, that is, the so-called "Ground Truth" of the device (that is, the data that enables the quality of the device model to be checked), are difficult to popularize in industry. Due to the high compatibility of the devices and the high costs caused by production interruptions, it is often also difficult to collect the "Ground Truth" for a production line. This method belongs to a method that is also referred to as "non intrusive load monitoring" in English, abbreviated as NILM. WO 2012145099 Al describes an example of this method, which gives a textual overview and points out the complexity of the traditional NILM method, which hinders its practical application. The NILM method claimed there is also based on the model of a single energy consumption monitoring device.

[0005] There is a need to perform low-overhead energy consumption determination in an industrial environment. SUMMARY OF THE INVENTION

[0006] The technical problem to be solved by the present invention makes a contribution thereto.

[0007] The above technical problem is solved by a method according to the present invention, an apparatus according to the present invention, a system according to the present invention, a computer program according to the present invention, and a computer program product according to the present invention. Advantageous expansion schemes are given in the following description.

[0008] According to the present invention, a new method for determining process-related energy consumption information of a plurality of automated processes is proposed. The automated processes can in particular be industrial processes. In the method according to the present invention, measurement data related to energy consumption of a plurality of processes is obtained. For this purpose, for example, data related to energy consumption of a plurality of processes within a certain time period is measured and the data related to energy consumption is transmitted for determining process-related energy consumption information. The measurement data is, for example, a current value and, if necessary, a voltage value, for example, a current value and, if necessary, a voltage value in the form of a measurement sequence covering the measurement time period. The transmitted measurement data can be processed data, for example, a consumption value calculated from the current and voltage values. Thus, the time period is preferably selected such that the time period at least includes the total duration of at least one work process or work step of all repeated work processes of all processes among the plurality of processes.

[0009] Then, the trend of energy consumption (for example, a trend curve) of a plurality of processes is determined by means of the measurement data related to energy consumption. Here, the determined energy consumption relates to all of the plurality of processes, or the measurement data related to energy consumption relates to the total consumption. Here, the term "energy consumption" should be interpreted as relating to a parameter related to energy consumption or energy efficiency (for example, active power, reactive power, total power,...).

[0010] In another step, process flow information of each process among the plurality of processes is obtained. This can be carried out at least in part by using the process flow information of a programmable controller, which is designed to control the processes among the plurality of processes.

[0011] Alternatively or additionally, the provision of process flow information can be provided, which is, for example, a video recording of each process or a group of processes among the plurality of processes, and is specifically generated for the method according to the present invention. Here, generating the process flow information can include identifying process activities and registering the associated time information (for example, a time stamp of the start or end of a process step).

[0012] In another step, the energy consumption change trends of multiple processes are matched with the process flow information of each process. Here, this matching of the energy consumption development of multiple processes with the process flow information of each process can include making the following curves consistent in time, where these curves describe the energy consumption development or the process flow of each process.

[0013] In another step, the changes in the energy consumption of multiple processes are associated with each process based on the process flow information. This can be done by identifying the changes in energy consumption in the energy consumption change trend. This identification of the changes can be carried out by means of the following criteria, through which it is ensured that the identified changes in energy consumption correspond to the relevant process flow changes (for example, threshold criteria for the magnitude or duration of the changes). This identification of the changes in energy consumption can include dividing the changes into groups that can be respectively associated with individual processes. Then, for example, it can be assumed that the same changes can be quantitatively associated with the same process. For example, artificial intelligence trained for this purpose can also be used to divide the groups.

[0014] Finally, the energy consumption information of each process is determined based on the changes in the energy consumption of multiple processes associated with the corresponding processes.

[0015] The present invention has the following advantages, namely, the relevant energy consumption information is simply determined based on the measurement data (such as current and voltage) collected for all processes. It is therefore low-cost in terms of the number of measurement components and in the determination (decomposition) of information related to individual processes.

[0016] The device designed to execute the method according to the present invention is also the subject of the present invention. It can be, for example, a computer to which the required measurement data is transmitted from the measurement device. This computer can also be responsible for executing the method according to the present invention for the entire pipeline. The device is thus preferably (but not necessarily) part of the pipeline. In this case, the measurement device of the pipeline can be designed to transmit the relevant measurement information to the device wirelessly or via a cable connection, where the transmission can be direct or through an intermediate node.

[0017] The system including the device according to the present invention and at least one measurement device, where the measurement device is designed to transmit measurement information to the device, is also the subject of the present invention. The system can include at least one programmable controller, and thus the system is configured to transmit the process flow information from at least one programmable controller to the device, or the device includes a programmable controller. In a pipeline with multiple programmable controllers, one of the controllers can also be configured as the device according to the present invention. Then, the other programmable controllers are configured to transmit the process flow information to the programmable controller serving as the device according to the present invention.

[0018] The system according to the invention may also include components for generating process flow information (such as at least one video recording device, possibly having a computing unit for processing the recorded video information) for performing the method according to the invention. It may also be provided to process or preprocess the generated process flow information by the components for generating process flow information and then transmit it to the device according to the invention.

[0019] A computer program and a computer program product having such a computer program (such as a data carrier) are also the subject of the invention, which, when executed on a computer, performs the method according to the invention.

[0020] The computer program is configured to be executable in at least one control unit. The computer program may be configured as software, such as an App (application program) configured to be downloadable from the Internet, or may be configured as firmware that can be stored in a memory and executed by a processor or a computing station. Alternatively or additionally, the computer program may also be at least partially configured as a fixed-wired circuit, such as may be configured as an ASIC (ASIC = Application Specific Integrated Circuit). Description of the Drawings

[0021] Hereinafter, the present invention will be described in detail within the scope of embodiments with reference to the drawings. In the drawings:

[0022] Figure 1a and Figure 1b Two alternative design solutions of the method according to the invention are shown,

[0023] Figure 2 A schematic illustration of a pipeline is shown,

[0024] Figure 3 showing, for example, as components corresponding to Figure 2 a schematic example of a high-performance multi-process machine (such as an injection molding machine) that is part of the pipeline,

[0025] Figure 4 showing Figure 3 the output data of the energy consumption measurement of the high-performance multi-process machine according to

[0026] Figure 5 and the detection of energy consumption change events for the output data of the energy consumption measurement, and Figure 4 showing the correlation of process flow data and energy consumption data for the data according to

[0027] Figure 6 Shown by Figure 5 A graphical representation of the associated energy consumption values ​​is obtained by decomposing the DETAILED DESCRIPTION

[0028] Figure 1a and Figure 1b A conceptual illustration of two different variants of the method according to the invention is shown.

[0029] The core of the design according to the invention is to match process flow information and energy consumption information to each other in order to decompose the energy consumption information. The process flow information can already exist or be generated in a manner specifically for the method according to the invention. It is also conceivable to combine these two possibilities, i.e. to supplement the existing process flow information with process flow information generated in a manner specifically for decomposition, in order to perform the decomposition as efficiently and completely as possible.

[0030] Figure 1a A first variant is shown. Here, it is assumed that process flow data already exists (generally in the programmable controller SPS), and the process flow data is accessed during the method (step 1). These process flow data involve multiple processes. Energy-related data (generally energy consumption curves describing power consumption recorded or determined by measuring equipment) are collected for these processes, which present the changing trend of energy consumption caused by the process flow of the process (step 21). In the next step (step 22), the changes in energy consumption (energy consumption change events) are identified. An important consideration of the present invention is that these changes can be related to the process flow events of the process (generally the start and end of the processing steps of the process). Therefore, in the next step (step 3), the energy consumption change is matched with the process flow by means of the energy consumption change event, or the energy consumption change is made consistent with the process flow in time. Here, the core is to relate the energy consumption change trend to the process information of multiple processes, thereby through the consistency in time, the energy consumption change event can be associated with a specific process. The quantitative change in energy consumption in the change process in the corresponding energy consumption change event thus corresponds to the effect of the associated process on the energy consumption (e.g., the quantitative change is thus a measure of the energy consumption of the process or of the additional energy consumption of multiple processes running simultaneously). The energy consumption data can thus be decomposed (consumption parts are associated with individual processes) (step 4).

[0031] Figure 1bVariants that involve the complete or partial absence of process flow data. In this case, process flow information is specifically generated according to the method of the present invention. This can be done, for example, by video recording of the process flow or the device executing the process flow (step 11), and then process flow data is generated from this video recording (step 12). For this purpose, the recording is analyzed, for example, by identifying the start and end of each process step and setting markers or timestamps for the corresponding time points. It thus corresponds to a change, that is, change events can be used to construct the time flow, and these change events can be associated with specific processes. That is to say, process flow data that can be related to the energy consumption change trend of the process is thus obtained.

[0032] Furthermore, the method according to the second case differs in step 23, in which the energy consumption change information is grouped in order to associate the energy consumption change with the process flow data. This grouping corresponds to a division into groups that respectively belong to a process. As a criterion, for example, the quantitative value of the energy consumption change is used. This value may fluctuate slightly for different energy consumption changes belonging to the same process. Here, "slightly" should be understood as small compared to the differences in energy consumption changes of different processes. In principle, the threshold of these slight fluctuations can be used to act on the grouping. However, in practice, it is meaningful to act using machine learning or artificial intelligence, that is, to train the corresponding KI programs (such as t-SNE, k-NN, Decision Tree, Random Forest, SVM, etc.) for grouping for complex situations.

[0033] The following is based on Figures 2 to 6 illustrates a specific method.

[0034] Figure 2 A pipeline is schematically shown. The pipeline is formed by modules 1 to module n, and workpieces are processed, for example, by means of these modules. Here, these modules can be individual machines or groups of machines, and multiple processes are respectively executed by these machines or groups of machines. These modules are determined (or the unit "module" is defined) in such a way that each module has its own programmable controller SPS, which includes the process flow of the process flow data belonging to the corresponding module. In addition, for each module, there is a measuring device, and the energy consumption of the module or relevant energy-related data (such as active power, reactive power) is collected through this measuring device. The measuring device can be, for example, a so-called PMD (power measurement device), which measures current and voltage, thereby determining the power value and transmitting this power value to a central location.

[0035] In Figure 3is shown according to Figure 2 a section of the pipeline. Here, the module is a high-performance multi-process machine, such as an injection molding machine, whose process is controlled by a programmable controller SPS. The programmable controller SPS is connected to a measuring device that collects the energy consumption of the machine or relevant energy-related data. Multiple processes are performed on the workpiece by the machine, or multiple processes for manufacturing the workpiece are performed by the machine.

[0036] The basic method according to the present invention will be illustrated below with the aid of a simple example. For this purpose, it is assumed that Figure 3 the machine only performs two processes. The energy consumption data measured by Figure 3 the measuring device is presented by the Figure 4 upper curve. As a next step, the changes ( Figure 4 lower curve) are identified, i.e., the changes in energy consumption or change events.

[0037] In the next step, the curve obtained according to Figure 4 is correlated with the process flow data. This is shown in Figure 5 . The programmable controller SPS extracts the time-varying trend curves of the two processes. The upper part in Figure 5 shows two trend changes. The middle curve corresponds to the energy consumption curve of the machine according to Figure 4 , and there is no difference (total consumption) between the two processes. The curve is made temporally consistent with the process trend change curve (pattern matching), where it is utilized that the consumption change event is caused by a change in one of the two processes (generally a switching-on or switching-off event).

[0038] Figure 5 The lower curve shows the result. The consumption change events have been associated with the two processes, and thus the influence of the processes on the energy consumption (decomposition of the energy consumption) can be determined. For example, the consumptions of the two processes are linearly added, and then the consumptions of process 1 and process 2 can be directly determined accordingly. This is shown with the aid of Figure 6 , Figure 6 shows Figure 5 a section of the lower curve.

[0039] Thus, the consumption E1 of process 1 and the consumption E2 of process 2 can be obtained.

[0040] In the method according to the present invention, if necessary, it can be taken into account that the situation may not be as ideal as in Figures 3 to 6 , for example,

[0041] a) The consumptions of different processes in the module or machine are not necessarily additive. In this case, Figure 6The value E2 therein will correspond to the additional consumption when processes 1 and 2 are carried out simultaneously. When the two processes are active, the parameter E12 corresponding to the energy consumption will be important. When there are sections where the two processes are carried out separately, the corresponding consumption when only one process is active can thus be derived based on this. In the described example, three relevant and different parameters will be produced, namely the consumption of each process when carried out separately and the consumption when carried out simultaneously.

[0042] b) The energy consumption may fluctuate, especially when the machine is switched on or machine components are connected (start of power supply) (for example due to the so-called inrush current). For example, this transient oscillation behavior can be taken into account in the method when determining the consumption by decomposition by taking the value that is temporally after the transient oscillation process.

[0043] c) Finally, when defects in terms of the stability of the consumption will have a negative impact on the required accuracy of the determined value, an average value can also be determined (for example, averaging the consumption value E1 within the time range T1 and averaging the consumption value E12 within the time range T2).

[0044] The description in the example process illustrates the principle underlying the present invention. Real systems are usually significantly more complex. A person skilled in the art can write a software program based on the above description, which also implements the steps of the method according to the present invention for complex systems. In particular, machine learning methods can also be used here.

Claims

1. A method for determining process-related energy consumption information of a plurality of automated processes, the method comprising the following steps: - obtain energy consumption-related measurement data for multiple processes, - determine the energy consumption trends of a number of processes with the aid of the energy consumption-related measurement data obtained, - obtain process flow information for each of the multiple processes, -Match the energy consumption trends of multiple processes with the process information of each process, - Correlate changes in energy consumption of multiple processes with individual processes based on process flow information, and - determining energy consumption information for each process based on changes in energy consumption of the plurality of processes associated with the respective process.

2. The method according to claim 1, It is characterized in that The process flow information is obtained at least in part by utilizing process flow information of a programmable controller designed to control a process of a plurality of processes.

3. The method according to claim 1 or 2, It is characterized in that Matching the energy consumption profiles of the plurality of processes to the process information of the individual processes includes making the following curves consistent in time, which describe the energy consumption development or the process of the individual processes.

4. The method according to any one of the preceding claims, It is characterized in that In order to associate changes in energy consumption of a plurality of processes with the individual processes based on process flow information of the processes, changes in energy consumption are identified.

5. The method according to claim 4, It is characterized in that The detected changes in energy consumption are divided into groups which can each be associated with an individual process.

6. A method according to any one of the preceding claims, comprising: - measure data related to energy consumption of multiple processes over a certain period of time, and - transmitting said measurement data for determining energy consumption information related to the process.

7. A method according to any one of the preceding claims, comprising: - providing process flow information at least in part by generating said process flow information, and - Obtain process flow information provided.

8. The method according to claim 7, It is characterized in that The process sequence information is generated using video recordings of individual processes or groups of processes from the plurality of processes.

9. The method according to claim 8, It is characterized in that Generating process flow information includes identifying process activities and registering the associated time information. 10 . An apparatus, the apparatus being designed to carry out the method according to claim 1 .

11. A system, wherein: - comprises an apparatus according to claim 10 and at least one measuring device designed to transmit measurement information to the apparatus, and - is designed to carry out the method according to claim 6.

12. The system according to claim 11, The system comprises at least one programmable controller, in, The system is designed to transmit process sequence information from the at least one programmable controller to the device, or the device includes a programmable controller.

13. The system according to claim 11 or 12, wherein: - includes components for generating process flow information, and - is designed to carry out the method according to any one of claims 7 to 9.

14. The system according to claim 13, The system includes at least one video recording device.

15. A computer program having a program code, which, when executed on a computer, performs the method according to any one of claims 1 to 9.

16. A computer program product comprising a computer program according to claim 15.

Images