CN116070734A - Transformer fault prediction method, device, equipment and storage medium - Google Patents
Transformer fault prediction method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN116070734A CN116070734A CN202211447332.5A CN202211447332A CN116070734A CN 116070734 A CN116070734 A CN 116070734A CN 202211447332 A CN202211447332 A CN 202211447332A CN 116070734 A CN116070734 A CN 116070734A
- Authority
- CN
- China
- Prior art keywords
- transformer
- target transformer
- working
- current
- condition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K13/00—Thermometers specially adapted for specific purposes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Theoretical Computer Science (AREA)
- Protection Of Transformers (AREA)
- Housings And Mounting Of Transformers (AREA)
Abstract
The invention discloses a transformer fault prediction method, a device, equipment and a storage medium, wherein the method comprises the following steps: performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain unknown parameters of the target transformer; according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer, a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer under the condition of large current is established; obtaining a degradation function of the target transformer according to a multi-state degradation model of the non-uniform continuous time hidden semi-Markov process; and predicting whether the iron core of the target transformer vibrates according to the degradation function. Therefore, a prediction model is built according to working data of the transformer, whether the iron core vibrates or not is predicted, and damage to the transformer is avoided.
Description
Technical Field
The invention belongs to the technical field of state evaluation of transformers, and particularly relates to a transformer fault prediction method, device and equipment and a storage medium.
Background
The transformer works under the condition of large current, whether the iron core of the transformer vibrates or not cannot be predicted in time, and the transformer is easy to damage. Therefore, how to predict whether the iron core of the transformer vibrates in time becomes the technical problem to be solved currently, the traditional mode mainly adopts expert assessment, the expert obtains information through own sense organs, and then judges whether the iron core of the transformer vibrates through own subjective consciousness and experience, but the expert assessment method is easily influenced by subjective factors such as personal experience and physical condition of the expert, so that the problem of inaccurate prediction is caused
Disclosure of Invention
The invention aims to solve the technical problems that: the transformer fault prediction method, device, equipment and storage medium solve the technical problem that whether the iron core of a transformer vibrates or not is difficult to predict accurately in the prior art, so that the transformer is damaged.
The technical scheme of the invention is as follows:
the invention provides a transformer fault prediction method, which comprises the following steps:
acquiring working current, working voltage and working oil temperature of a target transformer under a high current condition;
performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain unknown parameters of the target transformer;
according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer, a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer under the condition of large current is established;
obtaining a degradation function of the target transformer according to the multi-state degradation model of the non-uniform continuous time hidden semi-Markov process;
and predicting whether the iron core of the target transformer vibrates according to the degradation function.
Optionally, the performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current to obtain the unknown parameters of the target transformer includes:
obtaining a historical sequence number according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current;
carrying out likelihood function probability product maximization according to the historical sequence number to obtain a characteristic parameter set;
performing iterative optimization on the characteristic parameter set to obtain an optimal value of the characteristic parameter;
and taking the optimal value of the characteristic parameter as an unknown parameter of the target transformer.
Optionally, the performing iterative optimization on the feature parameter set to obtain an optimal value of the feature parameter includes:
performing matrix processing on the characteristic parameter set to obtain characteristic parameters and an observation probability matrix corresponding to the degradation process of the target transformer;
re-estimating and updating the observation probability matrix to obtain a new observation probability matrix;
optimizing the characteristic parameters to obtain new characteristic parameters;
obtaining a new characteristic parameter set according to the new characteristic parameter and the new observation probability matrix;
and repeatedly executing the operation until the new characteristic parameter set is equal to the estimated value of the preset characteristic parameter set, and obtaining the optimal value of the characteristic parameter.
Optionally, the multi-state degradation model of the non-uniform continuous time hidden half markov process of the target transformer working under the condition of large current is built according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer, and the multi-state degradation model comprises the following steps:
obtaining a conversion rate function of the target transformer according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer;
combining a kernel function of a multi-state degradation model of the non-uniform continuous time hidden half Markov process with a conversion rate function of the target transformer to obtain a kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current;
and establishing the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current according to the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process.
Optionally, the combining the kernel function of the multi-state degradation model of the non-uniform continuous-time hidden half markov process with the conversion rate function of the target transformer to obtain the kernel expression of the multi-state degradation model of the non-uniform continuous-time hidden half markov process of the target transformer working under the condition of high current includes:
obtaining a state sequence according to a kernel function of a multi-state degradation model of the non-uniform continuous time hidden semi-Markov process;
and converting the state sequence through a conversion rate function to obtain a kernel expression of a kernel function of a multi-state degradation model of the heterogeneous continuous time hidden semi-Markov process.
Optionally, the predicting whether the core of the target transformer vibrates according to the degradation function includes:
inputting the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to the degradation function to obtain a degradation parameter set;
and selecting optimal degradation parameters from the degradation parameter set to predict according to the degradation parameter set, and determining whether the iron core of the target transformer vibrates.
Optionally, the acquiring the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current includes:
acquiring a parameter acquisition instruction;
and collecting working current, working voltage and working oil temperature of the target transformer when the target transformer operates under the condition of high current through the current sensor, the voltage sensor and the temperature sensor according to the parameter collecting instruction.
In addition, in order to achieve the above object, the present invention also proposes a transformer failure prediction apparatus, including:
the acquisition module is used for acquiring the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current;
the processing module is used for performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain unknown parameters of the target transformer;
the modeling module is used for establishing a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer working under the condition of large current according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer;
the processing module is further used for obtaining a degradation function of the target transformer according to the multi-state degradation model of the non-uniform continuous time hidden semi-Markov process;
and the prediction module is also used for predicting whether the iron core of the target transformer vibrates according to the degradation function.
In addition, in order to achieve the above object, the present invention also proposes a transformer failure prediction apparatus, the apparatus comprising: a memory, a processor, and a transformer fault prediction program stored on the memory and running on the processor, the transformer fault prediction program configured to implement the transformer fault prediction method as described above.
In addition, in order to achieve the above object, the present invention also proposes a storage medium having stored thereon a transformer failure prediction program which, when executed by a processor, implements the transformer failure prediction method as described above.
The invention has the beneficial effects that:
the invention carries out unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain the unknown parameters of the target transformer; according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer, a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer under the condition of large current is established; obtaining a degradation function of the target transformer according to a multi-state degradation model of the non-uniform continuous time hidden semi-Markov process; and predicting whether the iron core of the target transformer vibrates according to the degradation function. Therefore, a multi-state degradation model of a non-uniform continuous time hidden semi-Markov process is established according to working data of the transformer under the condition of high current, whether iron core vibration occurs is predicted, corresponding repair measures are timely taken, and damage to the transformer is avoided.
Drawings
FIG. 1 is a schematic diagram of a transformer failure prediction apparatus in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a transformer failure prediction method according to a first embodiment of the present invention;
FIG. 3 is a flowchart illustrating a transformer failure prediction method according to a second embodiment of the present invention;
FIG. 4 is a flowchart illustrating a transformer failure prediction method according to a third embodiment of the present invention;
fig. 5 is a schematic functional block diagram of a transformer failure prediction apparatus according to a first embodiment of the present invention.
Detailed Description
Referring to fig. 1, fig. 1 is a schematic structural diagram of a transformer fault prediction device in a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the transformer fault prediction apparatus may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display (Display), and the optional user interface 1003 may also include a standard wired interface, a wireless interface, and the wired interface for the user interface 1003 may be a USB interface in the present invention. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (Wi-Fi) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) Memory or a stable Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
Those skilled in the art will appreciate that the configuration shown in fig. 1 does not constitute a limitation of the transformer fault prediction apparatus, and may include more or fewer components than shown, or may combine certain components, or may be a different arrangement of components.
As shown in fig. 1, an operating system, a network communication module, a user interface module, and a transformer failure prediction program may be included in a memory 1005, which is considered a type of computer storage medium.
In the transformer failure prediction apparatus shown in fig. 1, the network interface 1004 is mainly used for connecting to a background server, and performing data communication with the background server; the user interface 1003 is mainly used for connecting user equipment; the transformer fault prediction apparatus calls a transformer fault prediction program stored in the memory 1005 through the processor 1001 and executes the transformer fault prediction method provided by the embodiment of the present invention.
Based on the above hardware structure, an embodiment of the transformer fault prediction method of the present invention is provided.
Referring to fig. 2, fig. 2 is a flowchart illustrating a first embodiment of a transformer failure prediction method according to the present invention.
In a first embodiment, the transformer fault prediction method includes the steps of:
step S10: and acquiring the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current.
It should be understood that the execution subject of the present embodiment is a transformer remaining life prediction apparatus having functions of data processing, data communication, program running, and the like.
In a specific implementation, the transformer remaining life prediction device sends a collection instruction to the current sensor, the voltage sensor and the temperature sensor, so that the current sensor, the voltage sensor and the temperature sensor begin to collect the working voltage, the working current and the working oil temperature of the target transformer after receiving the collection instruction. So that the relevant working data of the target transformer under the condition of high current can be obtained, and the method can be used for predicting whether the iron core of the transformer vibrates.
It should be noted that, the temperature sensor used herein is measured by means of oil temperature chromatographic detection. The main reasons for the vibration of the transformer core are: the transformer is severely overloaded (overload operation); the input voltage exceeds more than 10% of the rated voltage; the transformer is incorrect in the installation process and is not firmly fixed; the design of the transformer itself is problematic; the load parameter matching of the transformer is problematic; the production process of the transformer has problems, and thus, it can be seen that, in a high current scene where the transformer is operated, the situation that the iron core vibrates easily causes the damage of the transformer.
Step S20: and performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain the unknown parameters of the target transformer.
It should be understood that, according to the available types of history information, the estimated methods may be classified into an unsupervised estimation method and a supervised estimation method, and in general, if the state prediction information and the real health state are not identified, the unsupervised estimation method is adopted, and in the present invention, the state prediction information and the real health state are not identified, so that the unsupervised estimation is adopted in the present invention.
In specific implementation, according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current, a historical observation sequence number is obtained; likelihood function probability product maximization is performed according to the historical sequence number, obtaining a characteristic parameter set; performing iterative optimization on the characteristic parameter set to obtain an optimal value of the characteristic parameter; and taking the optimal value of the characteristic parameter as an unknown parameter of the target transformer. Therefore, the historical observation sequence number can be obtained according to the relevant working data of the target transformer, and then the unknown parameters of the target transformer can be obtained according to the historical observation sequence number, so that a more accurate model of whether the iron core of the transformer vibrates or not can be constructed, whether the iron core vibrates or not can be predicted, corresponding repair measures can be timely taken, and damage to the transformer is avoided.
Step S30: and establishing a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer working under the condition of large current according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer.
It will be appreciated that the multi-state degradation model of the non-uniform continuous-time hidden semi-markov process employs a more general assumption of transformers operating under high current conditions for multi-state devices that are not observable state modeling.
In specific implementation, according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer, obtaining a conversion rate function of the target transformer; combining a kernel function of a multi-state degradation model of the non-uniform continuous time hidden half Markov process with a conversion rate function of the target transformer to obtain a kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current; and establishing the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current according to the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process. Therefore, a prediction model of whether the target transformer vibrates in the high-current iron core can be obtained, whether the iron core vibrates or not is predicted, repair measures for the target transformer are timely taken, and damage to the transformer is avoided.
Step S40: and obtaining a degradation function of the target transformer according to the multi-state degradation model of the non-uniform continuous time hidden semi-Markov process.
It should be understood that the degradation function should be able to output a prediction result of whether or not the core of the target transformer vibrates in the case of inputting the operating current and voltage and the operating temperature of the target transformer.
Step S50: and predicting whether the iron core of the target transformer vibrates according to the degradation function.
It should be noted that in the detection of conditions used for device diagnostics and prognostics, it is necessary to determine structures and parameters associated with the degradation and prediction process. Parameter estimation refers to the estimation of model parameters using historical observations, assuming that the number of degraded states (N) and possible state monitoring index values (v) are already known, then the other are unknown parameter sets that need to be estimated. The unknown parameters to be estimated are that the time between usage states can follow any arbitrary continuous time distribution definition.
In specific implementation, inputting the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to the degradation function to obtain a degradation parameter set; and selecting optimal degradation parameters from the degradation parameter set to predict according to the degradation parameter set, and determining whether the iron core of the target transformer vibrates. Therefore, a multi-state degradation model of a non-uniform continuous time hidden half Markov process is established according to the working current, the working voltage and the working oil temperature of the target transformer under high power, whether the iron core vibrates is predicted, corresponding repair measures are timely taken, and transformer damage is avoided.
In the embodiment, working current, working voltage and working oil temperature of a target transformer under a large current condition are obtained, and unsupervised estimation is performed on the working current, the working voltage and the working oil temperature of the target transformer under the large current condition to obtain unknown parameters of the target transformer; according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer, a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer under the condition of large current is established; obtaining a degradation function of the target transformer according to a multi-state degradation model of the non-uniform continuous time hidden semi-Markov process; and predicting whether the iron core of the target transformer vibrates according to the degradation function. Therefore, a multi-state degradation model of a non-uniform continuous time hidden semi-Markov process is established according to working data of the transformer under the condition of high current, whether iron core vibration occurs is predicted, corresponding repair measures are timely taken, and damage to the transformer is avoided.
Referring to fig. 3, fig. 3 is a flowchart illustrating a second embodiment of the transformer fault prediction method according to the present invention, and the second embodiment of the transformer fault prediction method according to the present invention is proposed based on the first embodiment shown in fig. 2.
In a second embodiment, the step S20 includes:
step S201: and obtaining a historical sequence number according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current.
It should be noted that in a target transformer operating under high current conditions, the degradation process between two discrete unobservable states conforms to either a markov or a non-markov architecture. In continuous time Markov degradation, the transition between two states is dependent only on the two states that are involved in the transition and is independent of the transition process.
It will be appreciated that the degenerate transformation between states i and j is represented by a transformation rate function, where t is represented as the duration of state i and s is represented as the point in time at which the transformer enters state i, where (s+t) is the process time or the total age of the device, the number of coefficients needed to characterize each transformation rate function variable is determined from the type of degenerate process associated with the transformation, and the set of transformation rate functions that make up is the number of historical sequences.
Step S202: and maximizing likelihood function probability products according to the historical sequence number to obtain a characteristic parameter set.
It should be noted that, the feature parameter set refers to the number of states of the device, feature parameters corresponding to the degradation process, and a set of state monitoring device index spaces, and the degradation and observation process related to the device is considered, and depends on available information, where some parameters may be unknown.
Step S203: and carrying out iterative optimization on the characteristic parameter set to obtain an optimal value of the characteristic parameter.
It should be noted that the parameter to be estimated represents a random relationship between the health of the device and the course of observation, this relationship being represented by an observation probability matrix, the elements of this matrix being unknown parameters of the model, the matrix having N rows and M columns. It should be noted that, where the failure state is directly observed through the observation process, it is necessary to specify the observed value in order not to lose the versatility.
In specific implementation, the characteristic parameter set is processed to obtain characteristic parameters and an observation probability matrix corresponding to the degradation process of the target transformer; re-estimating and updating the observation probability matrix to obtain a new observation probability matrix; optimizing the characteristic parameters to obtain new characteristic parameters; obtaining a new characteristic parameter set according to the new characteristic parameter and the new observation probability matrix; and repeatedly executing the operation until the new characteristic parameter set is equal to the estimated value of the preset characteristic parameter set, and obtaining the optimal value of the characteristic parameter.
Step S204: and taking the optimal value of the characteristic parameter as an unknown parameter of the target transformer.
In this embodiment, according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current, a historical observation sequence number is obtained; carrying out likelihood function probability product maximization according to the historical sequence number to obtain a characteristic parameter set; performing iterative optimization on the characteristic parameter set to obtain an optimal value of the characteristic parameter; and taking the optimal value of the characteristic parameter as an unknown parameter of the target transformer. Therefore, the fault prediction model of the target transformer can be perfected according to the unknown parameters of the target transformer. Therefore, whether the target transformer core vibrates or not can be accurately predicted, and faults of the transformer are avoided.
Referring to fig. 4, fig. 4 is a schematic flow chart of a third embodiment of the transformer fault prediction method according to the present invention, and the third embodiment of the transformer fault prediction method according to the present invention is proposed based on the first embodiment shown in fig. 2.
In a third embodiment, the step S30 includes:
step S301: and obtaining a conversion rate function of the target transformer according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer.
It will be appreciated that the transition between states of a semi-markov process using the function of modeling the transition rate will use the transition rate function to transition the associated stochastic process between state i and state j, assuming the device is in state i at time u, the probability of it transitioning to state j within a preferentially small time interval (u, u+d), the transition rate of the process at time u, is denoted as degenerate transition rate λ, and may be represented as shown in equation 1.
λdu＝Pr{u≤T n+1 -T n ≤u+du∩X n+1 ' S (1)
As shown in the above formula, T n The time at the nth conversion is indicated, and this is indicated by the age of the target transformer, and u indicates the time at which it is present. Transformation to be verifiedThe duration between the current state and the working time (total age of the device) is independent.
Step S302: and combining the kernel function of the multi-state degradation model of the non-uniform continuous time hidden half Markov process with the conversion rate function of the target transformer to obtain the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer working under the condition of high current.
It should be noted that the conversion function may be used to represent the transition between two states, and that the multi-state degradation model for a non-uniform continuous-time hidden semi-markov process has an important measure in terms of transition rate.
In a specific implementation, a state sequence is obtained according to a kernel function of a multi-state degradation model of a non-uniform continuous time hidden semi-Markov process; and converting the state sequence through a conversion rate function to obtain a kernel expression of a kernel function of a multi-state degradation model of the heterogeneous continuous time hidden semi-Markov process. Therefore, the kernel expression can be obtained, and a fault prediction model of the target transformer working under the condition of high current can be established, so that whether the target transformer vibrates or not can be obtained, measures can be timely taken, and the target transformer is prevented from being broken down.
Step S303: and establishing the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current according to the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process.
Let v= { V1,..v 2} be an observation space having M possible values, and define a random variable U. Wherein U is n Is the output observed during the nth conversion. The relationship between the degradation process and the observation process can be expressed as shown in equation 1.
Pr(U n ＝v j U C ,X C )＝Pr(U n ＝v j Xn＝i)＝b i (j) (2)
It should be understood that where Yi is defined as the output of the observation process at time t, where if Z is assumed to follow a non-uniform continuous-time semi-markov process, the (Z, Y) process is a non-uniform continuous-time hidden semi-markov process that will act as the target transformer to see if the core is vibrating in high current conditions, causing transformer failure.
In this embodiment, the conversion rate function of the target transformer is obtained according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer; combining a kernel function of a multi-state degradation model of the non-uniform continuous time hidden half Markov process with a conversion rate function of the target transformer to obtain a kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current; and establishing the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current according to the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process. Therefore, a fault prediction model of the target transformer under the condition of large current can be obtained, whether the target transformer generates iron core vibration or not can be accurately predicted, and the fault of the transformer can be accurately predicted.
In addition, the embodiment of the invention also provides a storage medium, wherein the storage medium is stored with a transformer fault prediction program, and the transformer fault prediction program realizes the steps of the transformer fault prediction method when being executed by a processor.
The technical solutions of all the embodiments can be adopted by the storage medium, so that the storage medium has at least the beneficial effects brought by the technical solutions of the embodiments, and the description is omitted herein.
Referring to fig. 5, fig. 5 is a schematic functional block diagram of a transformer failure prediction apparatus according to a first embodiment of the present invention.
In a first embodiment of the transformer failure prediction apparatus of the present invention, the transformer failure prediction apparatus includes:
the acquisition module 10 is configured to acquire an operating current, an operating voltage, and an operating oil temperature of the target transformer under a high current condition.
The processing module 20 is configured to perform unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current, so as to obtain an unknown parameter of the target transformer; and the method is also used for obtaining the degradation function of the target transformer according to the multi-state degradation model of the non-uniform continuous time hidden semi-Markov process.
The modeling module 30 is configured to build a multi-state degradation model of a non-uniform continuous-time hidden half markov process for the target transformer operating under a high current condition according to the working current, the working voltage, the working oil temperature and the unknown parameters of the target transformer under the high current condition.
The prediction module 40 is further configured to predict whether the iron core of the target transformer vibrates according to the degradation function.
In the embodiment, working current, working voltage and working oil temperature of a target transformer under a large current condition are obtained, and unsupervised estimation is performed on the working current, the working voltage and the working oil temperature of the target transformer under the large current condition to obtain unknown parameters of the target transformer; according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer, a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer under the condition of large current is established; obtaining a degradation function of the target transformer according to a multi-state degradation model of the non-uniform continuous time hidden semi-Markov process; and predicting whether the iron core of the target transformer vibrates according to the degradation function. Therefore, a multi-state degradation model of a non-uniform continuous time hidden semi-Markov process is established according to working data of the transformer under the condition of high current, whether iron core vibration occurs is predicted, corresponding repair measures are timely taken, and damage to the transformer is avoided.
In an embodiment, the acquiring module 10 is further configured to acquire a parameter acquisition instruction;
and collecting working current, working voltage and working oil temperature of the target transformer when the target transformer operates under the condition of high current through the current sensor, the voltage sensor and the temperature sensor according to the parameter collecting instruction.
In an embodiment, the processing module 20 is further configured to obtain a historical sequence number according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current;
carrying out likelihood function probability product maximization according to the historical sequence number to obtain a characteristic parameter set;
performing iterative optimization on the characteristic parameter set to obtain an optimal value of the characteristic parameter;
and taking the optimal value of the characteristic parameter as an unknown parameter of the target transformer.
In an embodiment, the processing module 20 is further configured to perform matrix processing on the feature parameter set to obtain a feature parameter and an observation probability matrix corresponding to the degradation process of the target transformer;
re-estimating and updating the observation probability matrix to obtain a new observation probability matrix;
optimizing the characteristic parameters to obtain new characteristic parameters;
obtaining a new characteristic parameter set according to the new characteristic parameter and the new observation probability matrix;
and repeatedly executing the operation until the new characteristic parameter set is equal to the estimated value of the preset characteristic parameter set, and obtaining the optimal value of the characteristic parameter.
In an embodiment, the modeling module 30 is further configured to obtain a conversion rate function of the target transformer according to the working current, the working voltage, the working oil temperature and the unknown parameters of the target transformer under the condition of high current;
combining a kernel function of a multi-state degradation model of the non-uniform continuous time hidden half Markov process with a conversion rate function of the target transformer to obtain a kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current;
and establishing the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current according to the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process.
In one embodiment, the modeling module 30 is further configured to obtain a state sequence according to a kernel function of a multi-state degradation model of the non-uniform continuous-time hidden semi-markov process;
and converting the state sequence through a conversion rate function to obtain a kernel expression of a kernel function of a multi-state degradation model of the heterogeneous continuous time hidden semi-Markov process.
In an embodiment, the prediction module 40 is further configured to input the working current, the working voltage, and the working oil temperature of the target transformer under the condition of high current to the degradation function, so as to obtain a degradation parameter set;
and selecting optimal degradation parameters from the degradation parameter set to predict according to the degradation parameter set, and determining whether the iron core of the target transformer vibrates.
Other embodiments or specific implementation manners of the transformer fault prediction device of the present invention may refer to the above method embodiments, so at least have all the beneficial effects brought by the technical solutions of the above embodiments, and are not described herein again.
Claims (10)
1. A method for predicting a transformer failure, the method comprising the steps of:
acquiring working current, working voltage and working oil temperature of a target transformer under a high current condition;
performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain unknown parameters of the target transformer;
according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer, a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer under the condition of large current is established;
obtaining a degradation function of the target transformer according to the multi-state degradation model of the non-uniform continuous time hidden semi-Markov process;
and predicting whether the iron core of the target transformer vibrates according to the degradation function.
2. The method for predicting a transformer fault according to claim 1, wherein performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of high current to obtain the unknown parameters of the target transformer comprises:
obtaining a historical sequence number according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current;
carrying out likelihood function probability product maximization according to the historical sequence number to obtain a characteristic parameter set;
performing iterative optimization on the characteristic parameter set to obtain an optimal value of the characteristic parameter;
and taking the optimal value of the characteristic parameter as an unknown parameter of the target transformer.
3. The method for predicting a transformer fault according to claim 2, wherein the performing iterative optimization on the feature parameter set to obtain the feature parameter optimal value includes:
performing matrix processing on the characteristic parameter set to obtain characteristic parameters and an observation probability matrix corresponding to the degradation process of the target transformer;
re-estimating and updating the observation probability matrix to obtain a new observation probability matrix;
optimizing the characteristic parameters to obtain new characteristic parameters;
obtaining a new characteristic parameter set according to the new characteristic parameter and the new observation probability matrix;
and repeatedly executing the operation until the new characteristic parameter set is equal to the estimated value of the preset characteristic parameter set, and obtaining the optimal value of the characteristic parameter.
4. The method according to claim 1, wherein the step of establishing a multi-state degradation model of a non-uniform continuous-time hidden half markov process for a target transformer operating under a high current condition based on an operating current, an operating voltage, and an operating oil temperature of the target transformer under the high current condition and unknown parameters of the target transformer comprises:
obtaining a conversion rate function of the target transformer according to the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current and the unknown parameters of the target transformer;
combining a kernel function of a multi-state degradation model of the non-uniform continuous time hidden half Markov process with a conversion rate function of the target transformer to obtain a kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current;
and establishing the multi-state degradation model of the non-uniform continuous time hidden half Markov process of the target transformer under the condition of high current according to the kernel expression of the multi-state degradation model of the non-uniform continuous time hidden half Markov process.
5. The method of claim 1, wherein the combining the kernel function of the multi-state degradation model of the non-uniform continuous-time hidden semi-markov process with the conversion function of the target transformer results in a kernel expression of the multi-state degradation model of the non-uniform continuous-time hidden semi-markov process for the target transformer operating under high current conditions, comprising:
obtaining a state sequence according to a kernel function of a multi-state degradation model of the non-uniform continuous time hidden semi-Markov process;
and converting the state sequence through a conversion rate function to obtain a kernel expression of a kernel function of a multi-state degradation model of the heterogeneous continuous time hidden semi-Markov process.
6. The method of claim 1, wherein predicting whether the core of the target transformer vibrates according to the degradation function comprises:
inputting the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to the degradation function to obtain a degradation parameter set;
and selecting optimal degradation parameters from the degradation parameter set to predict according to the degradation parameter set, and determining whether the iron core of the target transformer vibrates.
7. The method according to any one of claims 1 to 6, wherein the obtaining the operating current, the operating voltage, and the operating oil temperature of the target transformer in the case of a large current includes:
acquiring a parameter acquisition instruction;
and collecting working current, working voltage and working oil temperature of the target transformer when the target transformer operates under the condition of high current through the current sensor, the voltage sensor and the temperature sensor according to the parameter collecting instruction.
8. A transformer fault prediction apparatus, characterized in that the transformer fault prediction apparatus comprises:
the acquisition module is used for acquiring the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current;
the processing module is used for performing unsupervised estimation on the working current, the working voltage and the working oil temperature of the target transformer under the condition of large current to obtain unknown parameters of the target transformer;
the modeling module is used for establishing a multi-state degradation model of a non-uniform continuous time hidden half Markov process of the target transformer working under the condition of large current according to the working current and the working voltage of the target transformer under the condition of large current, the working oil temperature and the unknown parameters of the target transformer;
the processing module is further used for obtaining a degradation function of the target transformer according to the multi-state degradation model of the non-uniform continuous time hidden semi-Markov process;
and the prediction module is also used for predicting whether the iron core of the target transformer vibrates according to the degradation function.
9. A transformer fault prediction device comprising a memory, a processor and a transformer fault prediction program stored on the memory and executable on the processor, the transformer fault prediction program when executed by the processor implementing the transformer fault prediction method according to any one of claims 1 to 7.
10. A storage medium, wherein a transformer failure prediction program is stored on the storage medium, and wherein the transformer failure prediction program, when executed by a processor, implements the transformer failure prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211447332.5A CN116070734A (en) | 2022-11-18 | 2022-11-18 | Transformer fault prediction method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211447332.5A CN116070734A (en) | 2022-11-18 | 2022-11-18 | Transformer fault prediction method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116070734A true CN116070734A (en) | 2023-05-05 |
Family
ID=86169006
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211447332.5A Pending CN116070734A (en) | 2022-11-18 | 2022-11-18 | Transformer fault prediction method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116070734A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117074824A (en) * | 2023-08-17 | 2023-11-17 | 东莞市港龙电源设备有限公司 | Inspection system and method for transformer protection system |
-
2022
- 2022-11-18 CN CN202211447332.5A patent/CN116070734A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117074824A (en) * | 2023-08-17 | 2023-11-17 | 东莞市港龙电源设备有限公司 | Inspection system and method for transformer protection system |
| CN117074824B (en) * | 2023-08-17 | 2024-03-22 | 东莞市港龙电源设备有限公司 | Inspection system and method for transformer protection system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US12034319B2 (en) | Server-side characterization of rechargeable batteries | |
| US7107187B1 (en) | Method for modeling system performance | |
| US20230366936A1 (en) | Lithium ion battery lifetime prediction method, discharge capacity retention rate prediction method, lifetime prediction program, discharge capacity retention rate prediction program, and information processing device | |
| CN111459700A (en) | Method and apparatus for diagnosing device failure, diagnostic device, and storage medium | |
| CN111143173A (en) | Server fault monitoring method and system based on neural network | |
| CN112633421A (en) | Method and device for detecting abnormal electricity utilization behavior of user | |
| CN114371409B (en) | Training method of battery state prediction model, battery state prediction method and device | |
| CN111796233B (en) | Method for evaluating secondary errors of multiple voltage transformers in double-bus connection mode | |
| CN116413604A (en) | Battery parameter monitoring method, system, device and storage medium | |
| CN110766236A (en) | Power equipment state trend prediction method based on statistical analysis and deep learning | |
| CN116070734A (en) | Transformer fault prediction method, device, equipment and storage medium | |
| CN117131457B (en) | AI model-based electric power big data acquisition and processing method and system | |
| CN112561303B (en) | Power system dynamic analysis method based on integrated learning and power grid topology change | |
| CN114977176B (en) | Power load decomposition method, device, equipment and storage medium | |
| US20220243347A1 (en) | Determination method and determination apparatus for conversion efficiency of hydrogen production by wind-solar hybrid electrolysis of water | |
| Qian et al. | State of health estimation of lithium-ion battery using energy accumulation-based feature extraction and improved relevance vector regression | |
| CN112380759A (en) | Smart meter service life prediction method based on deep learning and CoxPH model | |
| CN115544803A (en) | Method, device, equipment and storage medium for predicting residual life of transformer | |
| CN116433034B (en) | Virtual power plant operation risk detection method based on machine learning | |
| CN116720983A (en) | Power supply equipment abnormality detection method and system based on big data analysis | |
| CN117195647A (en) | Method, apparatus, device, medium and program product for post-earthquake evaluation of transformer bushings | |
| CN112633583A (en) | Generator set vibration prediction method and device, computer equipment and storage medium | |
| CN111258863B (en) | Data anomaly detection method, device, server and computer readable storage medium | |
| CN111611117B (en) | Hard disk fault prediction method, device, equipment and computer readable storage medium | |
| CN112731150A (en) | Voltage sag state estimation method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |