• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    SYSTEM AND METHOD FOR POWER TRANSMISSION AND FORECAST AND DIAGNOSIS OF DISTRIBUTION RESOURCE CONDITION. The present invention relates to a computer-implemented system and method for predicting and diagnosing an electrical transmission, generation, and distribution resource health that includes a computer with a non-transient computer readout capable of receive data relating to a resource, its components, component subsystem and related parameters. Instructions stored on the non-transient, computer-readable medium execute instructions that predictively calculate the overall health of the resource and also calculate the states of subsystems and component parameters, providing a diagnosis of the causes of poor resource health.
    公开号:BR112015018578B1
    申请号:R112015018578-9
    申请日:2014-01-31
    公开日:2022-02-01
    发明作者:Luis Cheim；Lan Lin
    申请人:Abb Schweiz Ag；
    IPC主号:
    专利说明:

    Field of Invention
    [0001] The present description refers to the prediction and diagnosis of energy transmission conditions and distribution resources such as transformers, circuit breakers and batteries. In particular, the present disclosure relates to a statistical approach to evaluating false diagnosis, detection, and overall resource health. background
    [0002] Electrical resources such as transformers, circuit breakers and batteries involve tremendous cost and their proper maintenance is necessary to maximize the value realized from the resources by their life expectancy. A significant amount of operational information on resources can be obtained through sensors, supervisory and control systems, and through inspection by trained technicians. It may be desirable to analyze operational information in order to predict resource failures and to diagnose the causes of resource failures. summary
    [0003] A computer-implemented system and method for predictive analysis and diagnosis of an electrical transmission, generation and distribution resource health includes a computer with a computer-readable non-transient medium capable of receiving data relating to a resource, its components , component subsystem and related parameters. The instructions stored on the non-transient computer-readable medium execute the instructions that predictively calculate the overall health of the resource and also calculate the states of subsystems and component parameters, so as to provide a diagnosis of the causes of the poor health of the resource. Brief Description of Drawings
    [0004] In the attached drawings, structures and methods are illustrated which, together with the detailed description provided below, describe aspects of a system and method for diagnosing and predicting failures and the general condition of the electrical resource. It will be noted that a single component can be designed as multiple components or that multiple components can be designed as a single component.
    [0005] Furthermore, in the attached drawings and in the description below, like parts are indicated throughout the drawings and in the written description with the same reference numerals, respectively. The figures are not drawn to scale and the proportions of certain parts have been exaggerated for the sake of the better convenience of the illustration.
    [0006] Figure 1 illustrates a diagram of system 100.
    [0007] Figures 2-4 illustrate the Bayesian model 200 of resource health 102.
    [0008] Figure 5 illustrates a probability table 500 relating to the probability distribution of the states of the OPS random variable.
    [0009] Figure 6 illustrates method 600.
    [0010] Figure 7 illustrates method 700. Detailed Description
    [0011] Figure 1 illustrates a condition detection and diagnostics computer system 100 for use in managing a plurality of resources 102a to 102x of electrical power generation and distribution systems. Features 102a-102x illustrated in figure 1 are transformers. However, in accordance with other aspects of the present teachings the detection and diagnostic system 100 may be implemented with other features 102a-102x. The 102a-102x features may include, but are not limited to, various types of transformers such as large, medium, and small distribution transformers, power generation transformers, dry-type transformers, pole-mounted transformers, and high-energy transformers. . Features 102a-102x may also include non-transforming features, which include, but are not limited to, circuit breakers, switches, surge arresters, batteries, refrigeration systems, lines and line connections, relays, or other devices implemented in generation and energy distribution.
    [0012] The detection and diagnostics system 100 includes a computer 104. The computer 104 may include a processor 106, a non-transient computer-readable medium 108 such as a hard disk or random access memory (RAM) that can store executable instructions. 109, one or more databases 110a-110z and input data 111 such as provided by in-line sensors 118a-118y. The computer 104 also includes a supervisory control and data acquisition (SCADA) adapter 113 for communicating with the SCADA network 114 through connection 115 and communication adapters 116a-116y for communicating with in-line sensors 118a-118y through of communication connections 117a-117y. Communications between the internal components of the computer 104 may be affected via the bus 103.
    [0013] A user interface adapter 120 allows communication between the user interface 122 and the system 100 through the connection 121. The user interface 122 can take many forms, including but not limited to a touch screen, a keyboard, monitor or mouse. In other aspects, various forms of user interface 122 may be implemented at and connected to one or more suitable interface adapters 120. For example, configuration including a keyboard, monitor and mouse may be implemented with desktop computer 104. In another For example, system 100 may be implemented on a tablet device with a touch screen. In general and without limitation to the systems and methods described herein, it can be implemented in a variety of computing devices in a variety of forms including, but not limited to, mainframes, distributed systems, laptop computers, desktop computers and handheld devices such as like tablets.
    [0014] Communication adapters 116a-116y may have a variety of suitable shapes that allow data to be transmitted from inline sensors 118a-118y to computer 104. Connections 117a-117y between inline sensors 118a-118y and computer 104 can be hard-line connections or wireless connections, and can be in the form of a variety of communications protocols such as, for example and without limitation, DNP3.0, MODBUS or IEC61850. Standard internet or communications network protocols can also be implemented. As a non-limiting example only, one or more of the communication adapters 116a-116y can connect computer 104 to a company intranet or a proprietary TCP/IP network. Thus, any one or more of the in-line sensors 118a-118y can establish a connection 117a-117y through said company intranet or proprietary TCP/IP network. Inline sensors 118a-118y detect information input to sensor 119a-119y regarding the condition and operating parameters of resources 102a-102x, the condition and operating parameters of resource components 130, and the condition and operating parameters of subsystems of component 140a-140c. It should be noted that although resource 102a is shown with one resource component 130 and three subsystems 140a-140c, as discussed further herein a plurality of resource components 130 may be included in a particular resource 102a-102x, each with one or more more subsystems 140a-140c. Input information to sensor 119a-119y captured by in-line sensors 118a-118y results in an output by in-line sensors 118a-118y, which can be written as input data 111 on computer readable medium 108 and used to update one or more of the 110a-110z databases in accordance with the present teachings. As used herein, resource data 102, component data 130, component parameter data 130 and subsystem data refer to stored information relating to operational information regarding the condition of the respective resources 102 and the constituent parts and representative parameters, whether received via an in-line sensor 118a-118y, off-line sensor 124, SCADA network 114 or manually reported by readings from a technician or information from an expert.
    [0015] Inline sensors 118a-118y can provide output data related to resources 102a-102x to computer 104 continuously or intermittently. Intermittent signals may be provided to the computer 104 sporadically or may arrive after periodic time intervals ranging from milliseconds to days. Output data may be transmitted automatically by sensors 118a-118y or may be requested by computer 104.
    [0016] In addition to in-line sensors 118a-118y, off-line sensors 124 can also detect operational information 125 relating to resources 102. Data can be collected from off-line sensors 124 in a variety of ways. A technician may record a reading from sensor 124 and then manually input the data through user interface 122. Offline sensors 124 may also produce output from sensor 126 which may take the form of a reading from sensor 124, such as such as reading on a gauge located at sensor 124. In another example, data from sensor 124 may be written to a non-transient medium capable of being read by a local computer at sensor 124, which may then be transferred by doing the upload to computer 104, e.g. by copying to a portable non-transient medium which can be further copied to computer readable non-transient medium 108.
    [0017] In addition to data received from in-line sensors 118 and off-line sensors 124, operational data from resources 102a-102x can be obtained through the SCADA network 114 using remote terminal units (RTU) arranged locally on resources 102a-102x. Although the SCADA network 114 is shown in Figure 1, other data acquisition and resource control protocols may also be implemented in accordance with the present teachings, including but not limited to the Distributed Control System (DCS). Data received from the SCADA 114 network includes, but is not limited to, load magnitude and phase, current and voltage, ambient temperature, top oil temperature, winding temperature, relay actuation status, and various related alarms. to resources 102a-102x. SCADA data, for example, as received from local RTUs for resources 102a-102x, is transmitted over connections 128a-128x and is further transmitted to computer 104 over communication connection 115. Information received over the network of SCADA can be stored in a history that maintains long-term operational data about the particular resource 102a-102x. History may be stored as one or more databases 110a-110z on computer 104.
    [0018] In addition to the in-line sensors 118a-118y, the off-line sensors 124 and the information obtained from the SCADA network 114, other forms of resource assessments 102a-102x can be performed. For example, assessing a paint condition of features 102a-102x does not require instrumentation to detect, and can be performed by a technician who uses subjective judgment to assess the conditions. As discussed further herein, the assignment of a qualitative rating or a quantitative result may result from an on-line sensor reading 118a-118y or an off-line sensor reading 124 in one of the available states of a corresponding random variable. allows the modeling of that variable.
    [0019] In-line sensors 118a-118y, off-line sensors 124, information from the SCADA network 114 and any other evaluation performed on the 102a-102x resource may involve a large sampling of the operating conditions of the 102a-102x resources including the information received relative to resource 102a itself, as received via sensor 118c. More granular information is available by detecting parts of the 102a-102x resources. The condition or operating parameters of the resource component 130 is received by the sensor 118a, and the condition or operating parameters of the subsystems of the component 140b are received by the sensor 118b. Examples of 130 resource components include the load tap changer, oil conservation system, cooling system, bushings, surge arrester, main tank, and active part of 102a-102x transformer resources. Examples of 102a-102x resource parameters are operational history, maintenance log, and the number of failures experienced by 102a-102x resources. Parameters that are assignable to general resource 102a-102x can be identified as resource parameters. Examples of 140a-140c subsystems are the desiccant, which is a subsystem of the oil conservation system, or fans and pumps, which are a subsystem of the refrigeration system. Examples of component parameters include the number of operations as a parameter for the on-load tap-changer component 130, or the oil level as an oil conservation system parameter. It should be noted that certain parameters and subsystems 140a-140c will be shared or will be equally applicable on more than one component 130. For example, noise levels that cause harmful vibration may be shared between the main tank, the cooling system and the active.
    [0020] Referring to Figure 2, a Bayesian network 200 models the health condition of one of the resources 102a-102x, which may be referred to here individually or collectively as the resource 102. Reference to a resource 102 is an example and not limiting. The network 200 is in the form of a directed, acyclic graph having arrows 202 and nodes 204-288 representing the probability state of the resource, the parameters of the resource, the components 130, the subsystems of the component 140a-140c and the parameters of the component. . The probability distribution for a particular variable node is shown next to the node. Each node is associated with a random variable that corresponds to the resource's health, resource parameters, components 130, component subsystems 140a-140c, and component parameters. The random variable associated with any particular node may have one or more possible states or outcomes based on its particular characteristics. The probability distribution across said states or results reflects historical data, any specialized information incorporated into the distributions, and any updates to the distributions as may be performed in accordance with the teachings herein. Arrow 204 indicates that the status of resource parameters, components 130, component subsystems 140a-140c, and component parameters associated with the node from which the arrow originates has a causal relationship with resource 102 and the components 130 associated with nodes at which the arrow ends at the arrow head.
    [0021] Resource node 204 marked HINDEX represents the overall health of the system, as indicated by the value in the BOM result. Relatively high points in the HINDEX variable GOOD result indicate relatively better health of resource 102. A distribution of more than 50% BAD is considered a failure reading, which can trigger a technician to take action. It should be noted that the probability distributions at the nodes in figure 2 are normalized, so the probability that a random variable is in one of the available states of any particular node is 100%. According to one aspect of the present teaching, the distribution of results for each node can be from 2 to any integer number of states, whether the 2 conditions, for example, and without limitation fail and not fail, true and false, on and off or other state pairings.
    [0022] The LTC 206 component node corresponds to the on-load tap-changer, a mechanism that changes the number of active turns in a transformer winding while the transformer, such as resource 102, is in operation. The on-load switch has a variety of properties and subsystems, the corresponding states effecting the distribution of the state of the LTC node 206 as shown by the relationship dictated by the arrows 202 on the network 200.
    [0023] Various subsystems and parameters have a causal effect on the LTC 206 component node. The 208 subsystem node for a ContWear random variable corresponds to the wear of the tap-changer contacts that come into contact with the tap-changer. The condition of the contacts can be determined by visual inspection, the results of which can then be stored in the database 110a-110z. In-line sensors 118a-118y and off-line sensors 124 can also detect behavior associated with contact wear and performance and thereby infer contact wear and performance. The LTC_DGA 210 node corresponds to the random variable analysis of the dissolved gas analysis for the on-load tap-changer oil system. The on-load tap-changer oil system is separate from and occupies a smaller volume than the oil system for transformer resource 102. As such, the oil that occupies the on-load tap-changer oil system may degrade in a different way from the oil in the oil system of resource 102. Various in-line sensors 118a-118y and off-line sensors 124 are available to perform detection of gas dissolved in on-load tap-changer oil and providing empirical data corresponding to the LTC_DGA random variable. The OilFilter 212 node corresponds to the condition of the oil filter of a load tap-changer, which can be measured by the time since the last replacement or, for example, by checking the oil pressure in the filter. The ControlCabinet 214 node refers to the condition of a load tap-changer control cabinet. Evaluation of the control cabinet can be performed visually by determining whether any circuit connections are loose and whether any heaters installed in the cabin are operating properly. Mechanism node 216 corresponds to the mechanism status of a load tap changer, such as any motors, springs, switches, shafts, gears, and motor protection equipment. The mechanism evaluation can be performed visually, or it can be a function of the duration of service of the device 102. NumOp node 218 corresponds to the number of operations since the most recent previous maintenance was performed on a load tap-changer.
    [0024] The OPS 220 node corresponds to the oil conservation system of resource 102, which is responsible for separating the oil from the outside air and also conserves the oil by removing gases and contaminants from the oil. The subsystem and parameters having a causal effect on the oil conservation system include the desiccant, oil level and oil bag. Desiccant node 222 corresponds to desiccant, which removes moisture and can have a predetermined life expectancy. As such, the random variable associated with the desiccant node 222 may correspond to the life expectancy of the desiccant. Oil level node 224 corresponds to the oil level in the transformer. The Bag Member node 226 corresponds to the oil bag, which if compromised can result not in oil leakage but in contaminants within the oil.
    [0025] Cooling node 228 corresponds to the cooling system of resource 102. Node 230 which corresponds to the TopOilTemp random variable refers to the temperature of the oil inside the cooling system, while the FanPumps node 232 corresponds to the fans and pumps used to circulate the air around the heat exchanger devices of the resource 102 and facilitate the circulation of the oil within the resource 102. The current to the fans and pumps can be indicative of engine failure, and thus can serve as an observable of the random variable to the FanPumps 232 node.
    [0026] The ActivePart 234 node corresponds to the active part of the transformer 102 resource, which includes the transformer windings and core, accessories thereto, and mechanical supports for the winding and core. Node 236 associated with random variable FRA corresponds to frequency response analysis of transformer resource 102. Frequency response analysis is used to detect mechanical motion or damage to the active part of the transformer resource, including the winding and core. The purpose of frequency response analysis is to determine whether a displacement has occurred, with age or as a result of a particular event, including but not limited to events such as remodeling, repair, accident, failure, or transportation. Standard procedures for performing Frequency Response Analysis measurements can be found in the IEEE Draft Guide for the Application and Interpretation of Frequency Response Analysis for Oil Immersed Transformers; PC57.149/D9.3, August 2012. The DFR 238 node corresponds to dielectric frequency response analysis. Dielectric frequency response analysis involves measuring the capacitance and dielectric loss of insulating material over the frequency spectrum. The WindTTR 240 node is associated with the winding transformer turns ratio test. Transformer turns ratio testing involves testing the output voltage at the load when an input voltage is applied to the transformer. The InsPF 242 node is associated with the insulation energy factor, which is measured when determining the dielectric loss leakage current of the transformer oil resource 102. The AGE 244 node corresponds to the age of the live part. The GasInOil 246 node depends on the random variable for the GasRate 248 node, the GasLevels node 250 and the DuvalT 252 node. The GasRate 248 node corresponds to the gas production coefficient for one or more gases dissolved in the oil of resource 102, including but not limited to Hydrogen, Methane, Ethylene, Ethane, Acetylene, Propane, Carbon Monoxide and Carbon Dioxide. Node GasLevels 250 corresponds to the levels of gases dissolved in the oil of resource 102. The variable DuvalT refers to the result of the analysis of the Duval Triangle for resource 102, with the different possible state of the random variable corresponding to the different zones of the Triangle of Duval. The weighted distribution of the random variable states for the GasInOil 246 node are dependent on the random variable distributions for the GasRate 248 node, the GasLevels node 250 and the DuvalT 252 node. It should be noted that the GasRate 248 node, the GasLevels 250 node and the DuvalT 252 node have more than two possible outcomes.
    [0027] Bushings node 254 corresponds to bushings of resource 102. BushCPF 256 node and BshOilLev 258 node correspond to the capacitance and energy factor of the bushings, and the level of the insulation oil in the bushing, respectively. The MainTank node 260 matches the condition of the resource's main tank. The PaintFnsh 262 node corresponds to the state of the tank's external finish. For example, paint deterioration and exposure of the tank's underlying metal can increase the amount of corrosion in the tank and accelerate the onset of corrosion.
    [0028] Several other nodes represent the random variables associated with the subsystems and the component parameters the state of which affects the multiple components. For these nodes, the corresponding probability distributions have been omitted for the sake of clarity. Load node 264 corresponds to load on resource 102. OilLeak node 266 corresponds to the total number of leakage events or alternatively the amount of oil leakage. The OilQuality Node 268 corresponds to the presence or absence of contaminants in the insulating oil. The PDTest 270 node corresponds to partial discharge tests. During partial discharge tests a high voltage source is applied to the resource, and partial transformer discharges are performed and the resulting characteristics are observed. Partial discharge tests include dielectric frequency response (DFR) and frequency response analysis (FRA). The MainCab node 272 corresponds to the main control cabin of the resource 102 and its condition, including whether its heating device is operational. The InfraRed 274 node matches the thermal items detected by the infrared camera, for example, if any component has exceeded the reference temperature by a certain number of degrees. The HotSpot 276 node corresponds to the hotspot windings, which can be calculated from the oil temperature, load current, and data configuration of resource 102. The Calibers 278 node corresponds to the condition of the gauges in resource 102, such as like the temperature of the gauges and the oil level of the gauges in the main tank. NoiseVib 280 node corresponds to the physical vibration level of resource 102.
    [0029] Several resource and component parameters are shown for which there are no nodes having a causal relationship with said resource and component parameters so that the arrows point towards the resource parameters and components within the model 200, that is, so that no other node has the causal relationship with the nodes for said resource and component parameters. The ThruFault node 282 node reflects how many faults are experienced in a year, which is determined by how often a circuit breaker is tripped to protect resource 102. The Arrester node 284 corresponds to the over voltage protection system that, for example, minimizes lightning effects. The History 286 node corresponds to one or more of the maintenance, replacement, remodeling and failure history. The MAINT 288 node reflects recently performed maintenance or testing cases. The Switching node 290 corresponds to the operation of the resource switches. The TripProtect 292 node corresponds to the history of alarms and trips.
    [0030] In figure 2, preliminary distributions for the random variable can be determined using historical information so that it can be stored in the database 110. When determining the distributions, expert knowledge can also be fed to the system 100 such as by determining the thresholds to which outcomes correspond to particular states of a random variable, and the probability of failure conditions when certain likely conditions are met. Said preliminary probability distribution can serve as the previous distribution in the Bayesian analysis.
    [0031] Referring to Figure 3, the network 200 is being updated to reflect received data regarding the level of wear on the contacts of a load tap-changer. A technician who evaluates the contacts can, for example, inform the referred data. In Figure 3, the data that the contact failed completely sets the ContWear random variable at node 208 to 100% in a failure condition, referred to as BAD at node 208. The probability of the LTC node 206 being in a failure state is dependent due, in part, to the state of contact wear.
    [0032] In general, the notation P(A) corresponds to the probability of Where A is a parameter that can have one or more states, and P(A|B) denotes the probability of A considering B. In general, the theorem of Bayes is recited as: P(^|β) = P(BA)P(A)/P(B). Thus, the probability of A occurring considering B having occurred is equal to the probability of B occurring considering A times the probability of A occurring divided by the probability of B occurring. Bayes' theorem can be used to calculate an updated probability that the LTC 206 node is in the BAD or failed state by considering the determination that the state of the in-line switch contacts was in the BAD condition with Bayes' theorem: P (LTC = BADI ContWear = BAD) = (P(ContWear = BAD | LTC = BAD)P(LTC = BAD))/(P (ContWear = BAD)). Since the state probability of ContWear node 208 is reflected in node 208 as 100% BAD, LTC node 206 is adjusted to reflect the new probability distribution that a load tap changer is 82.01% likely to be in a fault condition, and 17.99% likely to be in a fault condition. Additionally, Bayes' theorem can be used to calculate the probability of resource 102 being in a fault condition: P(Asset = BADI LTC = BAD) = (P(LTC =BAD Asset = BAD)P(LTC = BAD) )/(P(Asset = BAD)). The result of the calculation is the index point shown in resource node 204 in figure 3, which shows that the updated health point for the resource is 66.90% unhealthy and 33.10% healthy.
    [0033] The updated distributions for the LTC node 206 and the HINDEX node 204 are determined by the system 100 based on the input data 111 stored on the non-transient computer-readable medium 108. In the example shown in Figure 3, the causal relationship is shown by the directed graph 200 between a load tap-changer contact wear represented by the ContWear node 208, and the condition of a load tap-changer represented by the LTC node 206, and in addition to the general resource represented by the HINDEX node 204. However, with reference to figure 4, based on data received regarding the system component can update the resource health information of the general resource 102a-102x, as well as the distribution of random variables that correspond to the subsystems and the parameters of the component conditions which incidentally affect the condition of the component.
    [0034] Still referring to figure 4, the data relating to the condition of a load tap-changer indicate the fault condition, so that with 100% certainty a load tap-changer has failed. The probability that the asset is in an unhealthy condition can be calculated by applying: P(Asset=BAD|LTC=BAD)=(P(LTC=BAD | Asset=BAD)P(LTC=BAD))/( P(Asset=R UIM)). With respect to subsystems and component parameters, the particular update value will be determined by applying Bayes' Theorem. For example, for any random variable X that incidentally affects a load tap changer, the probability that the random variable is in a BAD state can be determined by applying: P(X=BAD|LTC=BAD)=(P (LTC=BAD | X=BAD)P(LTC=BAD))/(P(X=BAD)). For example, where the random variable to update is ContWear, the probability that the random variable ContWear is in a negative state can be determined by applying: P(ContWear=BAD|LTC=BAD)=(P(LTC=BAD) | ContWear=BAD)P(L TC=BAD))/(P(ContWear=BAD)). In this way, diagnostic information can be determined with knowledge of failure of one of the components 130 or the resource 118a-118x itself.
    [0035] Referring to Figure 5 a portion of the distribution of the probability table 500 for the OPS node 220 that shows several possible states for the subsystems and the parameters 502 relevant to the oil conservation system of the resource 102a-102x. In particular, the portion of the table that shows the OilQuality, InfraRed, OilLeak, and Dessicant random variables having BAD conditions, by virtue of a greater than 50% probability of a BAD rating. The table shows the results when the results for the OilLevel node, the BagMemb node, and the Caliber node are GOOD or POOR, with the entries in table 504 corresponding to the health of the oil conservation system. Said distribution can be generated for all nodes 204-290 using historical data. Expert information can also be reported in the table, which represents the expert assessment of the various probability distributions. An expert can provide the initial relationship that is the source for the initial table of probabilities, including health condition or other initial distribution whether based on historical data or otherwise determined. Tables may be stored as data structures on computer readable media 108, for example, in a database 110 relating to the probability distribution of a particular node as a function of the random variables associated with the nodes that incidentally affect the particular node in the model 200. The table is also updated as Bayesian probabilities are calculated based on data received from sensors 118 or from tests by technicians, for example.
    [0036] Referring to Figure 6, the health prediction method 600 includes the step 602 of building a Bayesian model 200 of a resource 102. The Bayesian model 200 will be reflected at least in part in the instructions 109 stored on the readable medium by computer 108. In step 604, the prior probability distribution is created and stored in a database of prior probabilities of distributions, for example, by using the history of probability distributions and expert input information relating to the distributions of probability. In step 606, the subsystem or component parameter data is received, for example, as received from a sensor 118 and written as data 111 to computer readable medium that reflects the operating parameter information of a subsystem or component. In step 608, the probability distribution associated with the component node is updated using Bayes' Theorem, based on parameter data received from the subsystem or component. In step 610, the probability distribution associated with the resource node is updated using Bayes' Theorem, based on the updated probability distribution for the component node. In step 612, the previous probabilities created in step 604 are updated with the new probability distributions for the component node and resource node.
    [0037] Referring to Figure 7, the diagnostic method 700 includes the step 702 of building a Bayesian model 200 of a resource 102. As in Figure 6 and method 600, the Bayesian model 200 will be reflected at least in part in instructions 109 stored on computer readable medium 108. In step 704, the past probability distribution is created and stored in a database of past probability distributions, for example by using the history of probability distributions and the information entered by the expert concerning the probability distributions. In step 706, component data 130 is received, for example, as data 111 on computer-readable medium, which reflects operational information of a component 130. In step 708, the probability distribution associated with the nodes of the subsystem and component parameter nodes is updated using Bayes' theorem, based on the data received from the component. In step 710, the probability distribution associated with the resource node is updated using Bayes' theorem, based on the updated probability distribution for the component node. In step 712, the previous probabilities created in step 704 are updated with the new probability distributions for the resource node and nodes for the subsystems and component parameters.
    [0038] For purposes of the present description, unless otherwise specified, “the”, “a” or “a”, “an” means “one or more”. To the extent that the term "includes" or "including" is used in the specification or claims, it is intended to be inclusive in a similar way to the term "comprising" insofar as said term is interpreted when employed as a term of transition in claim. Additionally, to the extent that the term "or" is employed (e.g., A or B) it is intended to mean "A or B or both". When applicants intend to state “only A or B but not both” then the term “only A or B but not both” will be used. Thus, the use of the term "or" here is the inclusive use, not the exhaustive use. See, Bryan A. Garner, A Dictionary of Modern Legal Usage 624 (2d. Ed. 1995). Further, in the sense that the terms "in" or "within" are used in the specification or claims, it is intended that they additionally mean "in" or "on". As used herein, "about" will be understood by those skilled in the art and will vary to some extent depending on the context in which it is used. If there are uses of the terms that are not clear to those skilled in the art, considering the context in which it is used, "about" can mean up to plus or minus 10% of the particular term. From about A to B is intended to mean from about A to about B, where A and B are the specified values.
    [0039] While the present description illustrates various embodiments, and although said embodiments have been described in some detail, it is not the intent of the applicant to restrict or in any way limit the scope of the claimed invention to said details. Additional advantages and modifications will be apparent to those skilled in the art. Therefore, the present invention, in its broadest aspects, is not limited to the specific details and illustrative examples shown and described. Thus, deviations can be made from said details without departing from the spirit and scope of the claimed invention. Furthermore, the foregoing embodiments are illustrative, and no single feature or element is essential for all possible combinations that may be claimed in the present or subsequent application.
    权利要求:
    Claims (11)
    [0001]
    1. System (100) for determining a health of an electrical transmission or distribution resource (102a - 102x) having one or more subsystems (140a - 140c), characterized in that it comprises: a computer (104) that has a non-transient computer readable medium (108) and configured to receive at least one of component parameter data and subsystem data for at least one component (130) and the one or more subsystems (140a - 140c) of the resource, respectively; a probability distribution for each resource (102a - 102x), the at least one component (130), component parameter data, and the subsystem data written (604) on the computer-readable medium (108); instructions written on the medium computer-readable non-transient (108) that upon execution (608) update the probability distribution for the at least one component based on received subsystem data and component parameter data; instructions written on the non-transient medium computer-readable river (108) that upon execution (610) update the probability distribution for the resource (140a - 140c) based on the updated probability distribution of the at least one component (130); and instructions written on the non-transient computer-readable medium (108) that upon execution (612) update the probability distribution for the component parameter data and subsystem data based on the updated probability distribution of the resource and the at least one component.
    [0002]
    2. System (100) according to claim 1, characterized in that the computer (104) includes a user input interface (122) and is configured to receive at least one of the component parameter data and the subsystem data from at least one of an online sensor (118a - 118y) and the user input interface (122).
    [0003]
    3. System (100), according to claim 1, characterized in that the probability distribution of the at least one component includes probability distributions of at least one of an oil conservation system, a load tap changer , a cooling system, an active part, bushings and a main tank of a transformer.
    [0004]
    4. System (100), according to claim 3, characterized in that the probability distribution further includes probability distributions for at least one of component parameter data and subsystem data that include at least one of a oil quality, an oil leak, a charge or an infrared parameter.
    [0005]
    5. System (100), according to claim 1, characterized in that the probability distribution further includes probability distributions for at least one of the component parameter data and the subsystem data that includes at least one of an oil quality, an oil leak, a charge, or an infrared parameter.
    [0006]
    6. System (100), according to claim 1, characterized in that the probability distribution of the at least one component is determined based on the probability distribution of the subsystem data and the component parameter data, and a data structure associating the probability distribution of the subsystem data and the component parameter data with the probability distribution of the at least one component.
    [0007]
    7. System (100), according to claim 1, characterized in that the instructions written on the non-transient computer-readable medium (108) that upon execution (608, 610) update the probability distribution of the resource and the probability distribution of the at least one component, and instructions written on the computer-readable non-transient medium that upon execution (612) update the probability distribution of the component parameter data and subsystem data, perform a Bayesian calculation when of execution.
    [0008]
    8. System (100), according to claim 1, characterized in that the probability distribution of the resource, the probability distribution of the at least one component and the probability distribution of the at least one component parameter data and of subsystem data written to computer-readable media are normalized.
    [0009]
    9. System (100), according to claim 1, characterized in that it additionally comprises: written instructions on the non-transient computer-readable medium that upon execution update the probability distribution of the resource based on the resource data received .
    [0010]
    10. Method (600) for determining a health of an electrical transmission or distribution resource (102a - 102x) having one or more subsystems (140a - 140c), characterized in that it comprises: providing a computer (104) having a computer-readable non-transient medium (108) and configured to receive at least one of component parameter data for at least one component (130) of the resource (102a - 102x) and subsystem data for one or more subsystems (140a - 140c ); receiving (606) at least one of the component parameter data and the subsystem data with the computer (104); and, executing the instructions (608) written on the non-transient computer-readable medium (108) that update a probability distribution of the at least one component (130) based on the component parameter data and subsystem data received; executing the instructions (610) written on the non-transient computer-readable medium (108) that updates a probability distribution of the resource based on the updated probability distribution of the at least one component; and execute instructions (612) written on the non-transient computer-readable medium (108) that updates a probability distribution for at least one between component parameter data and subsystem data based on the updated distribution probabilities for the resource, and for at least one component.
    [0011]
    11. Method (600), according to claim 10, characterized in that it additionally comprises: executing written instructions on the non-transient computer-readable medium that update the probability distribution of the resource and the probability distribution of the component with based on received subsystem data and component parameter data.
    类似技术:
    公开号 | 公开日 | 专利标题
    BR112015018578B1|2022-02-01|System and method for determining a health of an electrical transmission or distribution resource
    EP1085534B1|2006-11-02|Intelligent analysis system and method for fluid-filled electrical equipment
    US8898525B2|2014-11-25|Method and system for use in condition monitoring
    EP1085635A2|2001-03-21|Fluid-filled electrical equipment intelligent analysis system and method
    JP2019505803A|2019-02-28|System and method for monitoring and diagnosing transformer health
    Ghazali et al.2009|TNB experience in condition assessment and life management of distribution power transformers
    CN110907064A|2020-03-24|GIS disconnecting switch contact temperature prediction method and device and readable storage medium
    JP2020030125A|2020-02-27|Abnormality diagnosis device and abnormality diagnosis method
    JPWO2011077530A1|2013-05-02|Method for predicting the possibility of abnormality in oil-filled electrical equipment
    CN109615557A|2019-04-12|Transformer equipment inspection base
    Krieg et al.2000|Techniques and experience in on-line transformer condition monitoring and fault diagnosis in ElectraNet SA
    Gautam et al.2020|Identifying Transformer Oil Criticality Using Fuzzy Logic Approach
    KR101600698B1|2016-03-08|System and Method for Predicting Life of Power Transformer
    Poiss2016|Development of DGA indicator for estimating risk level of power transformers
    Tenbohlen et al.2003|Benefit of sensors for on-line monitoring systems for power transformers
    Sefidgaran et al.2010|Reliability model of power transformer with ONAN cooling
    Dongale2008|Equipment condition assessment and its importance in estimation and prediction of power system reliability
    CN111044858B|2020-10-20|Risk assessment system and method for extra-high voltage converter transformer
    KR20210053846A|2021-05-12|Facility health monitoring method by measuring the electric circuit constant inside the power facility in operation
    Motau2019|Optimisation Of Online Condition Monitoring To Predict Power Transformer Failures In Distribution Networks
    Ciontu et al.2012|Preventive maintenance optimization for power transformers in use
    Ballal et al.2013|Computational intelligence algorithm based condition monitoring system for power transformer
    Yan et al.2014|A method for fast mining abnormal state information of power equipment based on time series analysis
    KR20210092630A|2021-07-26|Asset management method for electric power equipment
    Singh et al.2017|Use of online monitoring and data optimisation drives condition assessment on service-aged generator step up | transformers
    同族专利:
    公开号 | 公开日
    EP2951655B1|2017-03-08|
    CN105074598A|2015-11-18|
    EP2951655A1|2015-12-09|
    CN105074598B|2019-01-04|
    CA2900036A1|2014-08-07|
    EP2951655B8|2017-05-03|
    US10001518B2|2018-06-19|
    WO2014121113A1|2014-08-07|
    CA2900036C|2020-11-24|
    ES2620541T3|2017-06-28|
    BR112015018578A2|2017-07-18|
    US20140222355A1|2014-08-07|
    BR112015018578A8|2021-06-01|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    JPS61218323A|1985-03-20|1986-09-27|Toshiba Corp|Fault identification|
    US6415276B1|1998-08-14|2002-07-02|University Of New Mexico|Bayesian belief networks for industrial processes|
    KR100941558B1|2001-05-29|2010-02-10|웨스팅하우스 일렉트릭 컴퍼니 엘엘씨|Health monitoring display system for a complex plant|
    US7250907B2|2003-06-30|2007-07-31|Microsoft Corporation|System and methods for determining the location dynamics of a portable computing device|
    CN1276281C|2004-03-25|2006-09-20|烽火通信科技股份有限公司|Estimation method of polarizing membrane dispersion caused system unuseable probability|
    BRPI0502320A|2005-06-21|2007-02-06|Siemens Ltda|system and method of centralized monitoring and control of the operating condition of power transformers comprised of different substations and monitoring center|
    BRPI0502384A|2005-06-21|2007-02-06|Siemens Ltda|system and method of monitoring and controlling the operating condition of a power transformer|
    US20070050302A1|2005-08-25|2007-03-01|Cheim Luiz Americo V|Method for calculating the economic profitability of power transformers and method for the optimization of the economic profitability of power transformers|
    US7565262B2|2006-10-05|2009-07-21|Siemens Corporate Research, Inc.|Bayesian sensor estimation for machine condition monitoring|
    ES2690323T3|2011-04-15|2018-11-20|Abb Schweiz Ag|Dynamic evaluation system for high voltage electrical components|US10488384B2|2013-11-06|2019-11-26|Electric Power Research Institute, Inc.|System and method for assessing power transformers|
    US10176279B2|2015-06-05|2019-01-08|Uptake Technologies, Inc.|Dynamic execution of predictive models and workflows|
    US10579750B2|2015-06-05|2020-03-03|Uptake Technologies, Inc.|Dynamic execution of predictive models|
    US10176032B2|2014-12-01|2019-01-08|Uptake Technologies, Inc.|Subsystem health score|
    US10254751B2|2015-06-05|2019-04-09|Uptake Technologies, Inc.|Local analytics at an asset|
    US10878385B2|2015-06-19|2020-12-29|Uptake Technologies, Inc.|Computer system and method for distributing execution of a predictive model|
    US10176070B2|2015-05-19|2019-01-08|Goodrich Corporation|System and method for diagnosing line replaceable unit failure including measuring a hydraulic pressure and deriving a marginal distribution for each LRU|
    EP3350717A4|2015-09-17|2019-04-17|Uptake Technologies, Inc.|Computer systems and methods for sharing asset-related information between data platforms over a network|
    WO2017100306A1|2015-12-07|2017-06-15|Uptake Technologies, Inc.|Local analytics device|
    US10510006B2|2016-03-09|2019-12-17|Uptake Technologies, Inc.|Handling of predictive models based on asset location|
    US10796235B2|2016-03-25|2020-10-06|Uptake Technologies, Inc.|Computer systems and methods for providing a visualization of asset event and signal data|
    US10333775B2|2016-06-03|2019-06-25|Uptake Technologies, Inc.|Facilitating the provisioning of a local analytics device|
    US20180024508A1|2016-07-25|2018-01-25|General Electric Company|System modeling, control and optimization|
    US10210037B2|2016-08-25|2019-02-19|Uptake Technologies, Inc.|Interface tool for asset fault analysis|
    US10474932B2|2016-09-01|2019-11-12|Uptake Technologies, Inc.|Detection of anomalies in multivariate data|
    US10228925B2|2016-12-19|2019-03-12|Uptake Technologies, Inc.|Systems, devices, and methods for deploying one or more artifacts to a deployment environment|
    US10579961B2|2017-01-26|2020-03-03|Uptake Technologies, Inc.|Method and system of identifying environment features for use in analyzing asset operation|
    EP3370073B1|2017-03-01|2020-04-29|ABB Power Grids Switzerland AG|Method and device for determining capacitive component parameters|
    US10671039B2|2017-05-03|2020-06-02|Uptake Technologies, Inc.|Computer system and method for predicting an abnormal event at a wind turbine in a cluster|
    EP3407151A1|2017-05-24|2018-11-28|Tata Consultancy Services Limited|Systems and methods for cognitive control of data acquisition for efficient fault diagnosis|
    US10255526B2|2017-06-09|2019-04-09|Uptake Technologies, Inc.|Computer system and method for classifying temporal patterns of change in images of an area|
    US11232371B2|2017-10-19|2022-01-25|Uptake Technologies, Inc.|Computer system and method for detecting anomalies in multivariate data|
    US10552246B1|2017-10-24|2020-02-04|Uptake Technologies, Inc.|Computer system and method for handling non-communicative assets|
    US10379982B2|2017-10-31|2019-08-13|Uptake Technologies, Inc.|Computer system and method for performing a virtual load test|
    US10635519B1|2017-11-30|2020-04-28|Uptake Technologies, Inc.|Systems and methods for detecting and remedying software anomalies|
    EP3514908B1|2018-01-22|2022-02-09|Hitachi Energy Switzerland AG|Methods and devices for condition classification of power network assets|
    US10815966B1|2018-02-01|2020-10-27|Uptake Technologies, Inc.|Computer system and method for determining an orientation of a wind turbine nacelle|
    US10169135B1|2018-03-02|2019-01-01|Uptake Technologies, Inc.|Computer system and method of detecting manufacturing network anomalies|
    US10554518B1|2018-03-02|2020-02-04|Uptake Technologies, Inc.|Computer system and method for evaluating health of nodes in a manufacturing network|
    US10635095B2|2018-04-24|2020-04-28|Uptake Technologies, Inc.|Computer system and method for creating a supervised failure model|
    US10860599B2|2018-06-11|2020-12-08|Uptake Technologies, Inc.|Tool for creating and deploying configurable pipelines|
    US11137382B2|2018-06-15|2021-10-05|Morgan Schaffer Ltd.|Apparatus and method for performing gas analysis using optical absorption spectroscopy, such as infraredand/or UV, and use thereof in apparatus and method for performing dissolved gas analysison a piece of electrical equipment|
    US10579932B1|2018-07-10|2020-03-03|Uptake Technologies, Inc.|Computer system and method for creating and deploying an anomaly detection model based on streaming data|
    US11119472B2|2018-09-28|2021-09-14|Uptake Technologies, Inc.|Computer system and method for evaluating an event prediction model|
    US11181894B2|2018-10-15|2021-11-23|Uptake Technologies, Inc.|Computer system and method of defining a set of anomaly thresholds for an anomaly detection model|
    US11030067B2|2019-01-29|2021-06-08|Uptake Technologies, Inc.|Computer system and method for presenting asset insights at a graphical user interface|
    US11208986B2|2019-06-27|2021-12-28|Uptake Technologies, Inc.|Computer system and method for detecting irregular yaw activity at a wind turbine|
    US10975841B2|2019-08-02|2021-04-13|Uptake Technologies, Inc.|Computer system and method for detecting rotor imbalance at a wind turbine|
    US11262739B2|2019-09-05|2022-03-01|Hitachi Energy Switzerland Ag|Artificial intelligence/machine learning driven assessment system for a community of electrical equipment users|
    DE102020120539A1|2020-08-04|2022-02-10|Maschinenfabrik Reinhausen Gmbh|Device for determining an error probability value for a transformer component and a system with such a device|
    US20220058497A1|2020-08-21|2022-02-24|Siemens Industry, Inc.|Systems and methods for fault diagnostics in building automation systems|
    法律状态:
    2017-12-26| B25A| Requested transfer of rights approved|Owner name: ABB SCHWEIZ AG (CH) |
    2018-01-30| B25L| Entry of change of name and/or headquarter and transfer of application, patent and certificate of addition of invention: publication cancelled|Owner name: ABB TECHNOLOGY AG (CH) |
    2018-02-06| B25C| Requirement related to requested transfer of rights|Owner name: ABB TECHNOLOGY AG (CH) |
    2018-06-05| B25B| Requested transfer of rights rejected|Owner name: ABB SCHWEIZ AG (CH) |
    2018-06-05| B25L| Entry of change of name and/or headquarter and transfer of application, patent and certificate of addition of invention: publication cancelled|Owner name: ABB SCHWEIZ AG (CH) |
    2018-06-12| B25A| Requested transfer of rights approved|Owner name: ABB SCHWEIZ AG (CH) |
    2018-07-03| B25L| Entry of change of name and/or headquarter and transfer of application, patent and certificate of addition of invention: publication cancelled|Owner name: ABB SCHWEIZ AG (CH) |
    2018-11-13| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|
    2020-01-21| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|
    2021-11-30| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|
    2022-02-01| B16A| Patent or certificate of addition of invention granted [chapter 16.1 patent gazette]|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 31/01/2014, OBSERVADAS AS CONDICOES LEGAIS. |
    优先权:
    申请号 | 申请日 | 专利标题
    US13/759,026|2013-02-04|
    US13/759,026|US10001518B2|2013-02-04|2013-02-04|System and method for power transmission and distribution asset condition prediction and diagnosis|
    PCT/US2014/014235|WO2014121113A1|2013-02-04|2014-01-31|System and method for power transmission and distribution asset condition prediction and diagnosis|
    [返回顶部]