INTEGRATION OF PHYSICAL SENSORS INTO A FRAMEWORK FOR DATA ASSIMILATION

Abstract

Method for a modeling system (100), wherein the system (100) comprises a sensor network (102), a core data assimilation analytics layer (106, 206) and a local data assimilation analytics layer (104, 208) and is tuned by means of feedback to assess a state of a physical system, wherein the method has: Generating a model prediction through the core data assimilation analytics layer (106, 206) using a calibrated model of a physical system; Monitoring of measured data of the physical system using a plurality of sensors (108 - 111) connected to a forwarding node (112), wherein a first sensor from the plurality of sensors (108 - 111) contains a logic module configured to determine an uncertainty quantification and to combine the uncertainty quantification with the model prediction to output a state of the physical system, wherein the plurality of sensors (108 - 111) and the forwarding node (112) are part of a sensor network (102); Setting, on the sensor network (102), of at least one of a type and a parameter of a filter algorithm of the sensor network (102), based on the optimization of the type and / or parameter of the filter algorithm on the core and local data assimilation analytics levels (104, 106, 206, 208).

Description

Technical area

[0001] The present invention relates to computer modeling and, more particularly, to modeling the state of a complex physical system. BACKGROUND

[0002] Data assimilation is a method that combines observational data with the output of a numerical model to estimate the state of a system. Data assimilation networks may contain sensors that observe aspects of a physical system, such as ocean currents. The sensors may be integrated into a model state of a numerical model of the physical system. The sensors may perform measurement, basic data processing, and data transmission. The model state is a representation of the variables that attempts to fully describe the actual physical state of the system. The model is used because the available number of observations may be orders of magnitude smaller than the number of values ​​required to specify the state of the physical system.State-of-the-art physical sensors typically have capabilities that include measurement, basic data processing, and data transmission. However, considerable effort is required to integrate noisy and redundant sensor observations with predictive systems used to monitor and manage complex physical systems (e.g., fluid systems such as lakes).

[0003] There is therefore a need in technology to solve the above problem.

[0004] Document US 2018 / 0 307 997 A1 describes a system comprising a processor configured to retrieve sensor data, wherein the sensor data comprises one or more signals transmitted from a plurality of sensors arranged in a power generation system. The processor is further configured to group the sensor data into a first bin of data based on a degradation measure. The processor is additionally configured to quantify an uncertainty in the one or more bins of data, wherein the uncertainty comprises a difference between the one or more bins of data and a corrected data deviation inference.

[0005] Document US 2007 / 0 203 681 A1 describes a system and method for identifying hydrocarbon-productive reservoir zones using a numerical Monte Carlo analysis to account for uncertainties in log measurements and petrophysical parameters. Random errors in instrument responses and core measurements, as well as reservoir heterogeneity, are incorporated into a Monte Carlo simulation to output suitable petrophysical models that calculate statistical distributions for saturation and permeability to define productive zones, unproductive zones, and zones of uncertain productivity.

[0006] Document WO 2017 / 106743 A1 describes energy-efficient, modularized systems and methods for quantifying local uncertainties and predicting outcomes in mobile devices. An exemplary system implementing the described technology includes, for example, a mobile device with at least one sensor and a processor in an energy-efficient architecture. The processor executes an algorithm for quantifying uncertainty (e.g., Bayesian inference) on the data collected by the sensor and characterizes the uncertainty (e.g., the full posterior distribution) around latent variables of interest.

[0007] Document US 5,570,300 A describes a sensor for providing a measurement and information about the validity of the measurement. The sensor includes a transducer for generating a data signal related to the value of a variable and a transmitter for receiving the data signal and generating output signals. The transmitter generates a first output signal related to the value of the variable. The transmitter also generates a second output signal based on a dynamic uncertainty analysis of the first output signal. SUMMARY

[0008] Viewed from a further aspect, the present invention provides a method for assessing a state of a physical system, the method comprising: generating a model prediction using a calibrated model of a physical system; and monitoring measured data of the physical system using a plurality of sensors connected to a relay node, wherein a first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output a state of the physical system.

[0009] Viewed from a further aspect, the present invention provides a system comprising: a calibrated model of a physical system configured to generate a model prediction; and a plurality of sensors monitoring measured data of the physical system and connected to a relay node, wherein a first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output a state of the physical system.

[0010] Viewed from a further aspect, the present invention provides a computer program product for assessing a state of a physical system, the computer program product comprising a computer-readable storage medium readable by a processing circuit and storing instructions for execution by the processing circuit to perform a method for performing the steps of the invention.

[0011] Viewed from a further aspect, the present invention provides a computer program stored on a computer-readable medium and loadable into the internal memory of a digital computer, comprising software code portions for performing the steps of the invention when the program is executed on a computer.

[0012] A computer program product for assessing a state of a physical system, the computer program product comprising: a computer-readable storage medium having computer-readable program code embodied therein, the computer-readable program code executable by one or more computer processors to: generate a model prediction using a calibrated model of a physical system; and monitor measured data of the physical system using a plurality of sensors connected to a relay node, wherein a first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output a state of the physical system.

[0013] A method for assessing a state of a physical system according to an embodiment of the present invention, the method comprising generating a model prediction using a calibrated model of a physical system and monitoring measured data of the physical system using a plurality of sensors connected to a relay node, wherein a first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification related to the measured data and combine the uncertainty quantification with the model prediction to output a state of the physical system.

[0014] According to another embodiment, a system includes a calibrated model of a physical system configured to generate a model prediction, and a plurality of sensors monitoring measured data of the physical system and connected to a relay node, wherein a first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output a state of the physical system.

[0015] A computer program product for assessing a state of a physical system according to another embodiment, wherein the computer program product comprises a computer-readable storage medium having computer-readable program code embodied therein, the computer-readable program code being executable by one or more computer processors to generate a model prediction using a calibrated model of a physical system and to monitor measured data of the physical system using a plurality of sensors connected to a forwarding node, wherein a first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output a state of the physical system. BRIEF DESCRIPTION OF THE DIFFERENT VIEWS OF THE DRAWINGS

[0016] Now, and by way of example only, the present invention will be described with reference to preferred embodiments shown in the following figures: Fig. 1 is a block diagram of another embodiment of a data processing system configured to monitor, model, and enable determination of a state of a physical system; Fig. 2 is a block diagram of another embodiment of a data processing system configured to monitor and enable determination of a state of a physical system at the sensor level; Fig. 3 is a flowchart of one embodiment of a method for performing aspects of data assimilation to model a state of a complex physical system using sensor network processes; Fig. 4 is a flowchart of one embodiment of a method for performing aspects of data assimilation to model a state of a complex physical system using local data assimilation analytics; Fig. 5 is a flowchart of one embodiment of a method for performing aspects of data assimilation to model a state of a complex physical system using core data assimilation analytics; Fig. 6 shows a block diagram of a data processing system according to an embodiment; and Fig. 7 further illustrates a server such as the server from Fig. 6 according to one embodiment. DETAILED DESCRIPTION

[0017] One embodiment of a system may use data assimilation to optimally combine available observations with a numerical model to perform predictive analysis of the system. The system may integrate sensors within a framework for predictive data assimilation by enabling advanced sensor capabilities. One embodiment of the system may automatically integrate noisy and redundant sensor observations with predictive systems used to monitor and manage complex physical systems, such as a fluid system in a lake.

[0018] Related methods may output a state of the physical system using a calibrated model of the physical system. A calibrated model may be used to generate a model prediction. The system includes a plurality of sensors connected to a relay node and used to monitor measured data of the physical system. A first sensor of the plurality of sensors includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output the state of the physical system.

[0019] In contrast to conventional data assimilation, the system can incorporate a combination of local and edge (sensor-based) analytics to create a better data assimilation method. Edge analytics can be used to estimate sensor uncertainty and the weight assigned to measured values ​​relative to a model estimate.

[0020] A specific system implementation can bridge the gap between mere data acquisition through sensors and predictive online modeling. Predictive modeling can be focused on decision support in the management of complex physical systems. The system can incorporate integrated sensors and predictive models within a data assimilation framework.

[0021] An embodiment of the system can estimate the uncertainties of measured data in real time with high accuracy, using knowledge of the raw sensor data and a data filtering process. The system can identify sensors and parameters that can be modified in a raw data filtering process based on the accuracy of corrected model predictions. An automatic calibration mechanism of the system can use feedback from the data assimilation framework at the core data analytics level to optimize the type and parameters of the data filtering process at the sensor level.

[0022] One embodiment of the system may model a state of a complex physical system (e.g., a cyclone, a lake system, or a forested area). The system may use a calibrated, coupled model (e.g., weather, hydrodynamic, and / or water runoff models) of the physical system. The system may include a sensor network, local data assimilation analytics, and core data assimilation analytics.

[0023] The sensor network of one embodiment may include a plurality of sensors connected to the system via one or two forwarding nodes. The sensors may be configured to perform data acquisition, filtering of data stream data, and transmitting data to a cloud or central data warehouse. In particular, the sensors may additionally be configured to perform uncertainty quantification and provide real-time reporting. The sensors may acquire feedback from local and core data assimilation analytics. The sensors may also perform automatic type selection and parameter adjustment of a filter mechanism using system feedback. The sensors may use a wide range of filter types for data stream data.

[0024] Local data assimilation analytics can include components of an edge architecture, such as gateways, a messaging interface, and a sensor data store. In particular, data assimilation analytics can be responsible for optimizing and otherwise refining the filtering process at the sensor level.

[0025] Core data assimilation analytics may include sensor observations that can be assimilated. For example, the sensor observations can be combined with model predictions using information about uncertainty at a sensor level. The uncertainty estimate can be used dynamically according to the results of model performance or after data assimilation analytics.

[0026] Parameters at the sensor level can refer to a filtering process for data stream data. Parameters at the core data analytics level can refer to covariance calculations used to assimilate available measured values. A heuristic process can be used to refine the parameters at both levels.

[0027] The system may include a calibrated, coupled model (e.g., hydrodynamics, water runoff, weather) related to a physical system, as well as a plurality of sensors connected to the physical system via a relay node. The first of the plurality of sensors may be configured to determine an uncertainty quantification. The uncertainty quantification may be reported in real time. The system may combine the sensor observations with model predictions using information on a sensor-level uncertainty.

[0028] An embodiment of the system may enable advanced analytical capabilities at the sensor level to intelligently integrate physical sensing, modeling, and data assimilation. The system may perform observation processing on the sensor side, synchronized with data processing in the data assimilation framework. The data processing may include processes that go beyond simple filtering of the sensor readings. For example, the data processing in the sensors may include synchronized filtering and data assimilation. The system may enable quality control on the sensor side, guided by the data assimilation framework. One or more sensors may be configured to perform signal quality diagnostics.According to feedback from the data assimilation framework, an embodiment of a sensor may perform measurement noise characterization and adaptive sensor polling. The sensor may further provide diagnostics aimed at improving the sensor mesh network.

[0029] Data filtering may include processing data at the sensor level to reduce noise. An illustrative embodiment of data filtering may include a low-pass filter.

[0030] Data assimilation may include a process that uses observations from sensors of the physical system to correct predictions of a numerical model of that physical system. Local sensor data assimilation analytics may include analytics located at the sensor level (e.g., edge or gateway level) to optimize a filtering process of a raw data stream by refining parameters, using feedback on prediction accuracy at the core data assimilation analytics level.

[0031] An edge gateway can include a component capable of not only controlling sensor data traffic, but also storing and processing the data in one action. One embodiment's action can be performed by local data assimilation analytics. Core data assimilation analytics can combine sensor data and metadata with model predictions to improve model results. Using filtered real-time sensor data as well as information about the uncertainty of filtered real-time sensor data, core data assimilation analytics can also improve the data assimilation process.

[0032] In a specific implementation, the data assimilation system may include a network of sensor probes deployed in a lake. The sensor probes may be integrated with local and core data analytics. Each probe in the embodiment frequently acquires data on multiple environmental parameters at a constant rate. The generated data streams may be filtered at the probe level to reduce noise. Precise measurements and information on the uncertainty of the probe sensor data may be provided by the probe sensors in real time. The data assimilation system may refine the filtering process to obtain the best possible information at the sensor level, thus assimilating the data and data uncertainty, and thereby gradually obtaining better model predictions.

[0033] Referring again to the above example, a sensor for a specific variable (e.g., temperature) may collect a stream of raw data at a high sampling frequency (e.g., 4 Hz). Due to local environmental conditions, the raw data readings may be relatively noisy. The sensor may be equipped with a library of filters to preprocess the raw data.

[0034] A filtered data point (e.g., the output of a moving average filter) may contain an associated time and time interval. The filter may have a parameter that represents the length of a moving window. The uncertainty of the reported measured values ​​can be estimated during the filtering process. Any other type of filter (e.g., an adaptive filter) and filter parameters may produce a different preprocessed pair.

[0035] This feature differs from conventional sensors in that they neither report this uncertainty nor receive feedback to select a different filter (and its parameters). The uncertainty can be significantly influenced by the filter parameters and their changes over time. The parameters can be continuously refined using feedback from core data assimilation analytics.

[0036] Field measurements can be mapped to the physical model of the lake with a predefined spatiotemporal resolution. In this example, data assimilation may include the use of an optimal interpolation scheme.

[0037] A calculation and update may include measurement covariance matrices for the measured values ​​and the model. The calculation process may use real-time uncertainty estimates. Assimilation processes may include the optimal combination of data and data uncertainties, model predictions, modeling uncertainties, and covariance estimates. The output of the data assimilation analysis can be evaluated for improving model accuracy.

[0038] An optimization process of one embodiment of the system may refine parameters at the core data assimilation level by adjusting parameters of the state error covariance calculations (e.g., measurement uncertainties). Another or the same embodiment may refine parameters at the local data assimilation level by providing feedback to refine the data filter parameters. The modeling system may use data assimilation to optimally combine available observations with a numerical model to perform a predictive analysis of the physical system. Associated methods may output a state of the physical system using a calibrated model of a physical system, wherein the calibrated model is used to generate a model prediction.The system includes a plurality of sensors connected to a relay node and used to monitor measured data of the physical system. The system can integrate sensors within a framework for assimilating predictive data by enabling advanced capabilities for the sensor. At least one of the sensors includes a logic module configured to determine an uncertainty quantification regarding the measured data and combine the uncertainty quantification with the model prediction to output the state of the physical system. One embodiment of the system can automatically integrate noisy and redundant sensor observations with predictive systems used to monitor and manage complex physical systems.

[0039] Fig. 1 is a block diagram of one embodiment of a system 100 configured to model a state of a complex physical system. The system may include a sensor network 102 having a plurality of sensors 108, local data assimilation analytics 104, and core data assimilation analytics 106 (denoted by dashed lines). The physical system may include natural forces present in physical systems, e.g., a geographic region, a water system, a gravity system, a thermal system, a wind system, or the like.

[0040] The sensor network 102 of one embodiment may include a plurality of sensors 108 connected to the system 100 via one or two forwarding nodes 112. The sensors 108 may be configured to perform data acquisition, filtering of data stream data, and transmitting data to a cloud or central data warehouse. The sensors 102 may additionally be configured to perform uncertainty quantification and may report results in real time. The sensors 108 may acquire feedback from local and core data assimilation analytics. The sensors 108 may also perform automatic type selection and parameter adjustment of a filter mechanism using system feedback. The sensors 108 may use a wide range of filter types for data stream data.

[0041] In operation, the sensor network 102 may perform processes that include acquiring environmental data by the physical sensors 108 to 111 at a constant rate. A filtering algorithm executed in the sensors 108 (or in a processor that exchanges data with the sensors 108) may process the data stream to generate data that is stored in a historical database 118. The type of filtering algorithm and the filtering parameters may be automatically optimized at the core and local data assimilation analytics 104, 106 level and set accordingly at the sensor level 102. Based on automatic analysis during the filtering process, the sensors 108 may monitor the measurement uncertainty (e.g., an accuracy of the measurement values) and provide feedback thereon.

[0042] The local data assimilation analytics 104 may include components of an edge architecture such as gateways, a messaging interface, and a sensor data store. The data assimilation analytics 104 may be responsible for optimizing and otherwise refining the sensor-level filtering process. As described in Fig. 1, the local data assimilation analytics 104 may include an edge gateway module 114, a sensor messaging interface 116, a sensor data store 118, and a data assimilation analytics module 120.

[0043] Filtered sensor data can be received at the edge gateway module 114 and combined with relevant metadata to provide the filtered measurements, uncertainty characteristics, and information about the filtering process. Data can be forwarded to the data assimilation analytics module 120 via the sensor messaging interface 116. Local or edge analytics in the data assimilation analytics module 120 can optimize the filtering process by refining parameters using feedback to a predictive accuracy level at the core data assimilation analytics level.

[0044] Core data assimilation analytics 106 may include sensor observations that can be assimilated. For example, the sensor observations may be combined with model predictions using information about uncertainty at a sensor level. The uncertainty estimate may be used dynamically according to the results of model performance or after data assimilation analytics.

[0045] Parameters at the sensor level can refer to a filtering process for data stream data. Parameters at the core data analytics level can refer to covariance calculations used to assimilate available measured values. A heuristic process can be used to refine the parameters at both levels.

[0046] As in the embodiment from Fig. 1, the core data assimilation analytics 106 may include an observation processing module 122, a data assimilation module 124, a physical model module 126, an influence and boundary condition module 128, an improved model results module 130, and a data assimilation optimization module 132.

[0047] The physical model module 126 may model a state of a complex physical system (e.g., a lake system). The system may use a calibrated, coupled model (e.g., hydrodynamic, water runoff, and / or weather models) of the physical system.

[0048] The data assimilation module 124 may receive historical and real-time sensor data from the sensor data module 118 of the local data assimilation analytics 104. The sensor data may be filtered and include an indication of its uncertainty. The data assimilation module 124 may additionally include model predictions. The model predictions may be provided by the physical model module 126 and the influence and constraints module 128.

[0049] The observation processing module 122 may provide filtered observations and information about observation errors to the data assimilation module 124. The data assimilation module 124 may provide data assimilation feedback to the observation processing module 122.

[0050] The data assimilation module 124 can provide optimally combined observations and model results to the physical model module 126. The physical model module 126 can provide model results and information about modeling errors to the data assimilation module 124.

[0051] The data assimilation module 124 can combine the sensor data and metadata with the model predictions to improve model results. The data assimilation module 124 can optimize results by refining the data assimilation parameters and by providing feedback to refine the filtering / uncertainty estimation processes (e.g., at the sensor level).

[0052] Fig. Figure 2 is a block diagram of another embodiment of a data processing system 200 configured to monitor and enable determination of a state of a physical system. All or some of the Fig. 2 and implemented at the sensor level, such as in a sensor 202. Thus, the sensor 202 of the system 200 may be connected to or otherwise exchange data with additional sensors 204, a core data assimilation analytics layer 206, and a local data assimilation analytics layer 208. For example, the sensor module 202 may be connected directly or via a plurality of nodes 210.

[0053] The sensor module 202 or the sensor may include a processor 212, an interface 214, and a memory 216. As in the embodiment of the system 200 of Fig. As shown in Figure 2, memory 216 may include stored and real-time environmental data 218, a filter algorithm module 220, and parameters 222 for use in analysis. Memory 216 may additionally include data stream data 224, historical data 226, an uncertainty measurement module 228, an analysis module 230, a real-time feedback module 232, and a monitoring module 234.

[0054] The system 200 may perform processes that include acquiring the environmental data 218 by the physical sensors 202, 204 at a constant rate. The filtering algorithm 218 performed in the sensors 202, 204 may process the data stream to generate data that is stored in a historical database 226. The type of the filtering algorithm 218 and the filtering parameters 222 may be automatically optimized at the core and local data assimilation analytics 206, 208 level and set accordingly at the sensor level. Based on automatic analysis during the filtering process, the sensors 202, 204 may monitor and provide feedback on the measurement uncertainty.

[0055] The Fig. 3 through 5 are flowcharts illustrating embodiments of methods 300, 400, 500 of performing aspects of data assimilation in a manner consistent with embodiments of the system. The methods may be performed by a system such as the illustrative system 100 of Fig. 1 can be performed to model a state of a complex physical system using sensor network processes, local data assimilation analytics, and core data assimilation analytics.

[0056] With particular reference to the sensor network processes of the flow chart from Fig. 3, the sensor network 102 may perform processes at 302 that include acquiring environmental data by the physical sensors 108 to 111 at a constant rate. A filtering algorithm executed in the sensors 108 to 111 (or in a processor communicating with the sensors 108 to 111) may process the data stream at 304 to generate data that is stored in a historical database 118.

[0057] In 306, the type of filter algorithm and the filtering parameters can be automatically optimized at the core and local data assimilation analytics level 104, 106 and set accordingly at the sensor level 102. Based on an automatic analysis during the filtering process, the sensors 108 to 111 can monitor the measurement uncertainty (e.g., an accuracy of the measurement values) in 308 and provide feedback thereon.

[0058] With particular reference to the sensor network processes of the embodiment of the Fig. 4, filtered sensor data may be received in 402 in the edge gateway module and combined with relevant metadata in 404 to provide the filtered measured values, uncertainty characteristics, and information about the filtering process.

[0059] Data may be forwarded to the data assimilation analytics module 120 via the sensor messaging interface 116 at 406. Local or edge analytics in the data assimilation analytics module 120 may optimize the filtering process by refining parameters at 408 using feedback to a predictive accuracy at the core data assimilation analytics level.

[0060] With particular reference to the core data assimilation analytics processes from Fig. 5, the data assimilation module 124 may receive historical and real-time sensor data from the sensor data module 118 of the local data assimilation analytics 104 at 502. The sensor data may be filtered at 504 and include an indication of its uncertainty. The data assimilation module 124 may additionally receive model predictions at 506. The model predictions may be provided by the physical model module 126 and the influence and constraint module 128.

[0061] The data assimilation module 124 may combine the sensor data and metadata with the model predictions at 508 to improve model results. The data assimilation module 124 may optimize results by refining the data assimilation parameters at 510 and by providing feedback to refine the filtering / uncertainty estimation processes (e.g., at the sensor level) at 512.

[0062] Fig. 6 illustrates another example of a data processing system according to one embodiment, as may be implemented using a networked environment. As shown, the data processing environment 600 includes a client computer 605, a web server 610, a server 615, an application 616, and an application server 620. The client computer 605 may be a physical system (e.g., a desktop, a laptop computer, a mobile device, etc.) or a virtual computing instance running in the cloud. The client computer 605 includes a web browser 607. A user may access data services through the web browser 607 over a network 625 (e.g., the Internet).

[0063] For example, a user may access a web service 612 executing on a web server 610. In one embodiment, the web service 612 provides a web interface to an application server 620 (e.g., running an application service 622). Specifically, the application service 622 provides a database 624. The database 624 may contain data presented to users in the web browser 607.

[0064] Fig. Figure 7 further illustrates a server 715 such as the server 615 of Fig. 6 according to one embodiment. The server 715 generally includes a processor 704 connected via a bus to a memory 706, a network interface unit 718, a storage 708, an input unit 721, and an output unit 724. The server 715 is generally controlled by an operating system. Examples of operating systems include the UNIX operating system, versions of the Microsoft Windows operating system, and distributions of the Linux operating system (UNIX is a registered trademark of the Open Group in the United States and other countries. Microsoft and Windows are trademarks of Microsoft Corporation in the United States and / or other countries. Linux is a registered trademark of Linus Torvalds in the United States and / or other countries). More generally, any operating system that supports the functionality disclosed herein may be used.Processor 704 may represent a single CPU, multiple CPUs, a single CPU with multiple processing cores, and the like. Similarly, memory 706 may be random access memory. Although memory 706 is shown as a single identity, it should be understood that memory 706 may comprise a plurality of modules and that memory 706 may exist in several grades, from high-speed registers and caches to slower but larger DRAM chips. Network interface unit 718 may be any type of network data transmission unit that enables navigation server 710 to communicate with other computers over network 525.

[0065] Memory 708 may be a persistent storage device. Although memory 708 is shown as a single device, memory 708 may be a combination of fixed and / or removable storage devices, such as hard disk drives, solid-state disks, removable memory cards, optical storage, and network storage systems.

[0066] As shown, memory 706 contains application 717, which may be an application generally executed to perform the actions described herein. Memory 708 contains algorithms 714, filtered measurements 717, and uncertainty characteristics 719.

[0067] The input unit 721 may provide a keyboard and / or a mouse, etc. The output unit 724 may be any conventional display screen. Although shown separately from the input unit 721, the output unit 724 may be combined with the input unit 721. For example, a display screen with an integrated touch-sensitive screen may be used.

[0068] The descriptions of the various embodiments of the present invention have been presented for illustrative purposes and are not intended to be exhaustive or limited to the disclosed embodiments. Those skilled in the art will appreciate that numerous modifications and variations are possible without departing from the spirit and scope of the described embodiments. The terminology used herein has been chosen to best explain the principles of the embodiments, practical application, or technical improvement over current technologies, or to enable others skilled in the art to understand the embodiments disclosed herein.

[0069] Reference is made below to embodiments presented in this disclosure. However, the scope of the present disclosure is not limited to specific embodiments described. Rather, any combination of the following features and elements, regardless of whether they relate to other embodiments, is contemplated to realize and implement this contemplated disclosure. Furthermore, although embodiments disclosed herein may achieve advantages over other possible solutions or over the prior art, whether a particular advantage is achieved by a given embodiment does not constitute a limitation on the scope of the present disclosure.The following aspects, features, embodiments, and advantages are therefore merely illustrative and are not to be considered elements or limitations of the appended claims unless expressly relied upon in one or more claims. Likewise, reference to "the invention" is not to be construed as a generalization of any inventive subject matter disclosed herein and is not to be considered an element or limitation of the appended claims unless expressly relied upon in one or more claims.

[0070] Aspects of the present invention may accordingly take the form of an embodiment implemented entirely in hardware, an embodiment implemented entirely in software (e.g., firmware, resident software, microcode, etc.), or an embodiment that combines software and hardware aspects, which may be collectively referred to as a "circuit," "module," or "system."

[0071] The present invention may be a system, a method, and / or a computer program product. The computer program product may include a computer-readable storage medium (or computer-readable storage media) having computer-readable program instructions stored thereon for causing a processor to perform aspects of the present invention.

[0072] The computer-readable storage medium may be any physical device capable of retaining and storing instructions for use by an instruction execution unit. The computer-readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. A non-exhaustive list of more specific examples of the computer-readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM).Flash memory), a static random access memory (SRAM), a portable CD-ROM, a DVD (Digital Versatile Disc), a memory stick, a floppy disk, a mechanically encoded device such as punched cards or raised structures in a groove on which instructions are stored, and any suitable combination thereof. A computer-readable storage medium, as used herein, shall not be construed as containing transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., pulses of light traveling through fiber optic cables), or electrical signals transmitted through a wire.

[0073] Computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to respective computing / processing units or to an external computer or storage unit via a network such as the Internet, a local area network (LAN), a wide area network (WAN), and / or a wireless network. The network may include copper transmission cables, fiber optic transmission lines, wireless transmission, routers, firewalls, switching units, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing unit receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium within the respective computing / processing unit.

[0074] Computer-readable program instructions for performing operations of the present invention may be assembly language instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, or the like, as well as conventional procedural programming languages ​​such as the C programming language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server.In the latter case, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, over the Internet using an Internet service provider). In some embodiments, electronic circuits, including, for example, programmable logic circuits, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), may execute the computer-readable program instructions by utilizing state information of the computer-readable program instructions to personalize the electronic circuits to perform aspects of the present invention.

[0075] Aspects of the present invention are described herein with reference to flowchart and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart and / or block diagram, as well as combinations of blocks in the flowchart and / or block diagram, may be implemented using computer-readable program instructions.

[0076] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine, such that instructions executing via the processor of the computer or other programmable data processing apparatus produce a means for implementing the functions / steps specified in the flowchart and / or block diagram block(s). These computer-readable program instructions may also be stored on a computer-readable storage medium capable of directing a computer, a programmable data processing apparatus, and / or other devices to function in a particular manner, such that the computer-readable storage medium on which instructions are stored comprises an article of manufacture, including instructions implementing aspects of the function(s) specified in the flowchart and / or block diagram block(s).implement the function / step specified in the blocks of the flow charts and / or block diagrams.

[0077] The computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of process steps to be performed on the computer, other programmable apparatus, or other device to produce a computer-implemented process such that the instructions executing on the computer, other programmable apparatus, or other device implement the functions / steps specified in the block(s) of flowcharts and / or block diagrams.

[0078] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowcharts or block diagrams may represent a module, segment, or portion of instructions comprising one or more executable instructions for performing the particular logical function or functions. In some alternative implementations, the functions specified in the block may occur in a different order than shown in the figures. For example, two blocks shown in succession may actually execute substantially concurrently, or the blocks may sometimes execute in reverse order depending on the corresponding functionality.It is further to be understood that each block of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by special purpose hardware-based systems that perform the specified functions or steps, or by combinations of special purpose hardware and computer instructions.

[0079] Embodiments of the invention may be provided to end users via a cloud computing infrastructure. Cloud computing generally refers to the provision of scalable computing resources as a service over a network. More formally, cloud computing can be described as a computing capability that provides an abstraction between the computing resource and its underlying technical architecture (e.g., servers, storage, networks), thus enabling convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal administrative effort or interaction with a service provider. Thus, cloud computing enables a user to access virtual computing resources (e.g.,storage, data, applications, and even entire virtualized computing systems) in “the cloud” without having to consider the underlying physical systems (or locations of those systems) used to provide the computing resources.

[0080] Cloud computing resources are typically provided to a user on a pay-per-use basis, where users are only charged for the computing resources actually used (e.g., an amount of storage space used by a user or a number of virtualized systems instantiated by the user). A user can access any of the resources located in the cloud at any time and from anywhere on the Internet. In connection with the present invention, a user can access applications (e.g., threshold adjustment algorithms) or associated data available in the cloud. For example, the modules of Fig.1 may be executed in a cloud computing system, and in such a case, the threshold adjustment algorithms may adjust response thresholds and store the new values ​​in a cloud storage location. This allows a user to access this information from any computing system connected to a network connected to the cloud (e.g., the Internet).

[0081] Although the foregoing explanations are directed to embodiments of the present invention, other and further embodiments of the invention may be developed without departing from the essential scope thereof. Although certain embodiments are applicable to spoken language systems, the claims are not limited to or particularly applicable to spoken language interfaces. In one example, an embodiment of a method may not relate to a speech modality. The essential scope is thus determined by the following claims.

Claims

[1] A method for a modeling system (100), the system (100) comprising a sensor network (102), a core data assimilation analytics layer (106, 206) and a local data assimilation analytics layer (104, 208) and being tuned by means of feedback to assess a state of a physical system, the method comprising: generating a model prediction by the core data assimilation analytics layer (106, 206) using a calibrated model of a physical system; Monitoring measured data of the physical system using a plurality of sensors (108 - 111) connected to a forwarding node (112), wherein a first sensor of the plurality of sensors (108 - 111) includes a logic module configured to determine an uncertainty quantification and combine the uncertainty quantification with the model prediction to output a state of the physical system, wherein the plurality of sensors (108 - 111) and the forwarding node (112) are part of a sensor network (102); Setting, at the sensor network (102), at least one of a type and a parameter of a filter algorithm of the sensor network (102) based on the optimization of the type and / or parameter of the filter algorithm at the core and local data assimilation analytics levels (104, 106, 206, 208). [2] The method of claim 1, further comprising reporting the uncertainty quantification in real time by the plurality of sensors (108 - 111). [3] The method of claim 1, wherein the uncertainty quantification is determined in real time using the set of raw data and the filtering process. [4] The method of claim 1, wherein the corrected model prediction is output. [5] The method of claim 4, further comprising modifying an execution in the first sensor based on an accuracy of the corrected model prediction. [6] The method of claim 4, wherein the parameter of the assimilation process refers to a covariance calculation used in assimilating the raw data. [7] The method of claim 1, wherein the physical system comprises at least one of the following natural forces: a geographical region, a water system, a gravitational system, a thermal system, and a wind system. [8] A feedback-tuned system for assessing a state of a physical system, comprising: a sensor network (102) comprising a relay node (112) connected to a plurality of sensors (108-111), including a first sensor configured to: Determining an uncertainty quantification of the first sensor that describes feedback on the accuracy of the measurements taken by the first sensor during a filtering process based on automatic analysis, wherein the plurality of sensors (108 - 111) and the forwarding node (112) are part of the sensor network (102); and Selecting at least one type and one parameter of a filtering algorithm of the filtering process at the first sensor of the sensor network (102) based on feedback from a core and local data assimilation analytics layer (104, 208); a core data assimilation analytics layer (106, 206) configured to: Generating a model prediction using a calibrated model of the physical system; Combining model prediction and uncertainty quantification to improve model results; and Providing feedback to the sensor network (102) to select at least one type and one parameter of a filter algorithm of the filtering process at the first sensor of the sensor network (102); and a local data assimilation analytics layer (104, 208) configured to: Providing feedback to the sensor network (102) to select at least one type and one parameter of a filtering algorithm of the filtering process at the first sensor of the sensor network (102). [9] The system of claim 8, wherein the plurality of sensors (108-111) are configured to provide a message in real time. [10] The system of claim 8, wherein the uncertainty quantification is determined using raw data and a real-time filtering process. [11] The system of claim 8, wherein the output state of the physical system is a corrected model prediction. [12] The system of claim 11, wherein an execution in the first sensor is modified based on an accuracy of the corrected model prediction. [13] The system of claim 11, wherein a parameter is modified based on an accuracy of a corrected model prediction, the parameter being a covariance calculation used in assimilating the measured data. [14] The system of claim 8, wherein the physical system comprises at least one of the following natural forces: a geographical region, a water system, a gravitational system, a thermal system, and a wind system. [15] The system of claim 8, wherein the output state is modified by refining data assimilation parameters. [16] A computer program product for assessing a state of a physical system, the computer program product comprising: a computer-readable storage medium readable by a processing circuit and storing instructions for execution by the processing circuit to perform a method according to any one of claims 1 to 7. [17] A computer program product stored on a computer-readable medium and loadable into the internal memory of a digital computer, comprising software code portions for carrying out the method according to any one of claims 1 to 7 when the program is executed on a computer.

Images