Fault diagnosis method and system for motor driving mechanism of high-voltage circuit breaker

Abstract

The invention discloses a fault diagnosis method and system for a motor driving mechanism of a high-voltage circuit breaker, relates to the technical field of high-voltage switch equipment, and comprises an intelligent fault diagnosis method combining interactive multi-model and unscented Kalman filtering. Unscented Kalman filtering is utilized, so that the nonlinear characteristic of a state equation under the fault condition can be better approximated, and the filter can have better stability and lower calculated amount requirements; by means of the interactive multi-model, the problem that fault diagnosis is difficult due to the fact that measurable parameters are few is solved, and the rapidity and accuracy of fault diagnosis under the condition that the difference between the generated fault and the pre-assumed fault is large are improved. In the diagnosis process, an active matching mechanism is used for improving the problem that the matching degree of a fault model and an actual fault is poor, faults are actively found or eliminated, and the accuracy and rapidity of fault diagnosis are improved.

Description

Technical Field

[0001] The invention relates to the technical field of high-voltage switchgear, and in particular to a method and system for diagnosing faults of a motor drive mechanism of a high-voltage circuit breaker. Background Art

[0002] Failure of the motor drive mechanism of a high-voltage circuit breaker will not only affect the accurate opening and closing of the moving contact of the high-voltage circuit breaker, but may also cause the high-voltage circuit breaker to trip. Sudden faults of components and sensors are random, and due to the limited measurement parameters and the influence of the sensor's own faults on the accuracy of fault diagnosis, it is a very challenging task to accurately and quickly diagnose sudden faults of the motor drive mechanism of a high-voltage circuit breaker. The multi-model adaptive estimation method is an intelligent fault diagnosis method composed of a Kalman filter and Bayesian reasoning. In this method, the Kalman filter is used to track the changes in the fault characteristic parameters of each fault model, and the Bayesian reasoning is used to obtain the probability of each model, and then the fault condition is judged by the probability size. The limitation of the multi-model adaptive estimation method is that when the fault is greatly different from the pre-assumed fault, that is, when the residual deviation is large, the Kalman filter cannot obtain satisfactory estimation results and is easily affected by nonlinearity and diverges. This paper proposes a multi-model adaptive estimation fault diagnosis method based on the combination of interactive multi-model and unscented Kalman filter. The unscented Kalman filter is used to replace the Kalman filter in the traditional multi-model adaptive estimation method, and the interactive multi-model is used to feed back the historical fault probability into the Bayesian reasoning, which improves the speed and accuracy of fault diagnosis when the fault that occurs is very different from the pre-assumed fault. Summary of the invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that when performing fault diagnosis of a high-voltage circuit breaker motor drive mechanism, the limitation of the multi-model adaptive estimation method is that when the fault that occurs is significantly different from the pre-assumed fault, that is, when the residual deviation is large, the Kalman filter cannot obtain a satisfactory estimation result and is easily affected by nonlinearity and diverges.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: A method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism, comprising: constructing a first model reflecting the high-voltage circuit breaker motor drive;

[0006] The parameter vector of sensor failure and the parameter vector of system component failure are respectively introduced into the first model to construct a second model reflecting the sensor failure of the motor drive mechanism of the high-voltage circuit breaker and a third model reflecting the system component failure of the motor drive mechanism of the high-voltage circuit breaker; the second model and the third model are interactively input to update the state; the second model and the third model after the state update are updated in time, measurement value and probability, and the updated states are merged to obtain a fourth model reflecting the failure of the motor drive mechanism of the high-voltage circuit breaker; an active matching mechanism is introduced into the fourth model, the model parameters are corrected, and the fault type is predicted based on the corrected model.

[0007] As a preferred solution of the fault diagnosis method of the motor drive mechanism of a high-voltage circuit breaker described in the present invention, the second model and the third model contain several sub-models, and the number of the sub-models is determined by the parameter vector, specifically including: the parameter vector of the sensor fault and the parameter vector of the system component fault contain multiple sub-vectors, after introducing a sub-vector in the parameter vector of the sensor fault into the first model, a sub-model in the second model reflecting the fault type corresponding to the current sub-vector is obtained, when all sub-vectors in the parameter vector of the sensor fault are introduced into the first model, all sub-models of the second model are obtained, and all sub-models of the second model together constitute the second model; after introducing a sub-vector in the parameter vector of the system component fault into the first model, a sub-model in the third model reflecting the fault type corresponding to the current sub-vector is obtained, when all sub-vectors in the parameter vector of the system component fault are introduced into the first model, all sub-models of the third model are obtained, and all sub-models of the third model together constitute the third model.

[0008] As a preferred solution of the high-voltage circuit breaker motor drive mechanism fault diagnosis method described in the present invention, the interactive input to the second model and the third model for state update includes updating the state estimation value and covariance of the second model and the third model.

[0009] As a preferred solution of the method for diagnosing faults of a high-voltage circuit breaker motor drive mechanism described in the present invention, before the time updating, measurement value updating and probability updating of the second model and the third model after the state update, it also includes using the first method to process the second model and the third model after the state update to obtain the estimated discrete nonlinear system state of each model; the time updating includes updating the estimated value and the error covariance on the time scale of each model based on the first method.

[0010] As a preferred solution of the method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism described in the present invention, wherein: the measurement value update includes, based on the first method, obtaining the predicted estimated value of the measurement equation of each model under the estimated discrete nonlinear system state, obtaining the gain according to the predicted estimated value, and updating the estimated value and error covariance at the scale of each model measurement equation based on the gain; the probability update includes, converting each model into a likelihood function, and using the second method to update the likelihood function to a probability model; the fusion of the updated state to obtain a fourth model reflecting the fault of the high-voltage circuit breaker motor drive mechanism includes, fusing the estimated value and error covariance at each scale of each model, and embedding them into the probability model to obtain a fourth model reflecting the fault of the high-voltage circuit breaker motor drive mechanism, and the fourth model has the same structure as the second model and the third model, and contains several fourth model sub-models.

[0011] As a preferred solution of the high-voltage circuit breaker motor drive mechanism fault diagnosis method described in the present invention, the active matching mechanism is introduced into the fourth model, and the model parameters are corrected, including using a fault parameter to represent each fault type, and the fault parameter is a numerical range. The fault parameter is substituted into the fourth model, and the fault probability of each fourth model sub-model is calculated, and the average probability of the fault probability is obtained. If the fault probability of any fourth model sub-model is greater than the average probability and higher than the threshold, the fault parameter in the current fourth model sub-model is corrected so that the fault probability of the current fourth model sub-model meets the threshold requirement.

[0012] As a preferred solution of the fault diagnosis method of the motor drive mechanism of a high-voltage circuit breaker described in the present invention, wherein: the introduction of an active matching mechanism into the fourth model and the correction of the model parameters also include keeping the fault parameters unchanged, calculating the fault probability of each fourth model sub-model at the next moment, and obtaining the average probability at the next moment, if the fault probability of the fourth model sub-model at any next moment is greater than the average probability at the next moment and higher than the threshold, and the fault probability of the fourth model sub-model at the next moment is greater than the fault probability of the fourth model sub-model at the previous moment, then the fault probability of the fourth model sub-model at the current next moment is corrected so that the fault probability of the fourth model sub-model at the current next moment meets the threshold requirement.

[0013] Another object of the present invention is to provide a high-voltage circuit breaker motor drive mechanism fault diagnosis system, which can perform fault diagnosis on the high-voltage circuit breaker motor drive mechanism.

[0014] To solve the above technical problems, the present invention provides the following technical solutions: a system for a method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism, comprising: a model building module, a state updating and fusion module and a correction module; the model building module builds a first model reflecting the high-voltage circuit breaker motor drive, introduces a parameter vector of a sensor fault and a parameter vector of a system component fault into the first model respectively, and builds a second model reflecting the sensor fault of the high-voltage circuit breaker motor drive mechanism, and a third model reflecting the system component fault of the high-voltage circuit breaker motor drive mechanism; the state updating and fusion module inputs the second model and the third model interactively to update the state, updates the time, measurement value and probability of the second model and the third model after the state update, and fuses the updated state to obtain a fourth model reflecting the fault of the high-voltage circuit breaker motor drive mechanism; the correction module introduces an active matching mechanism into the fourth model, corrects the model parameters, and predicts the fault type based on the corrected model.

[0015] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned method for diagnosing a fault of a motor drive mechanism of a high-voltage circuit breaker when executing the computer program.

[0016] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for diagnosing a fault of a motor drive mechanism of a high-voltage circuit breaker as described above.

[0017] The beneficial effects of the present invention are as follows: the present invention utilizes unscented Kalman filtering to not only better approximate the nonlinear characteristics of the state equation, but also enables the filter to have better stability and lower computational requirements. The limitation of the traditional method of using a multi-model adaptive estimation method when performing fault diagnosis of a high-voltage circuit breaker motor drive mechanism is that when the fault that occurs differs greatly from the pre-assumed fault, that is, when the residual deviation is large, the Kalman filter cannot obtain a satisfactory estimation result and is easily affected by nonlinearity and diverges. The present invention utilizes unscented Kalman filtering to be robust to parameter uncertainty, filter tuning and initial state, and is insensitive to noise and not easy to diverge; and the use of interactive multi-models not only solves the problem of difficulty in fault diagnosis caused by a small number of measurable parameters, but also improves the rapidity and accuracy of fault diagnosis when the fault that occurs differs greatly from the pre-assumed fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0019] Figure 1 This is a flow chart of a method for diagnosing faults of a high-voltage circuit breaker motor drive mechanism in Example 1.

[0020] Figure 2 This is a framework diagram of a multi-mode adaptive estimation fault diagnosis algorithm for a high-voltage circuit breaker motor drive mechanism fault diagnosis method in Example 1.

[0021] Figure 3 This is a flow chart of an active pattern matching algorithm for a method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism in Example 2.

[0022] Figure 4 This is a module structure diagram of a high-voltage circuit breaker motor drive mechanism fault diagnosis system in Example 4. DETAILED DESCRIPTION

[0023] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0024] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0025] Example 1, reference Figure 1-Figure 3 , which is the first embodiment of the present invention, and which provides a method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism, including: Figure 1 As shown:

[0026] S1. Construct a first model reflecting the motor drive of a high-voltage circuit breaker.

[0027] The first model construction method can establish corresponding differential equations, algebraic equations and state equations according to the physical characteristics of the motor drive mechanism. These models usually include motor body model, drive circuit model, load model, etc. Collect motor drive data under normal working conditions and fault conditions, including current, voltage, speed, temperature, etc.

[0028] The present invention describes the high-voltage circuit breaker motor drive mechanism model as the following differential equation and measurement equation,

[0029]

[0030] In the formula, is the system state vector at discrete time t k =kT s The sampling value, T s is the sampling period of the system; is the system control input vector; is the measured output vector; f(x(k),u(k)) is the set of nonlinear functions of the state vector and the control input vector; is the process noise vector; H(x(k)) is the measurement function; is the measurement noise vector, which is independent of the process noise vector w(k). It indicates that the value range of the variable is real numbers, and the superscript n, p, and m indicate the dimension of the variable.

[0031] The data collected in this model define the phase current I a ,I b ,I c , motor rotor position θ and speed ω m The state vector x includes the phase voltage V a ,V b ,V c and load torque T L The control input vector u includes the stator winding inductance L a and resistor R a 、Motor inertia J m and the damping coefficient B m The corresponding component fault injection parameter vector f a ; Including the motor rotor position θ and the high-voltage circuit breaker moving contact displacement s parameter input vector f s ; and by the phase current I a ,I b ,I c , the measurable output vector z composed of the motor rotor position θ and the high-voltage circuit breaker moving contact displacement s.

[0032] S2. Introduce the parameter vector of sensor failure and the parameter vector of system component failure into the first model respectively, and construct a second model reflecting the sensor failure of the high-voltage circuit breaker motor drive mechanism and a third model reflecting the system component failure of the high-voltage circuit breaker motor drive mechanism.

[0033] In order to detect faults through changes in model parameters, the established fault model must be sensitive to this type of fault, and fault models for sensing sensors (second model) and system component changes (third model) are established respectively. When establishing the second model and the third model, machine learning algorithms (such as neural networks, support vector machines, random forests, etc.) can be used to train feature parameters and build a fault diagnosis model.

[0034] In order to reduce the system burden, the present invention adopts the method of introducing fault vectors to construct the second model and the third model, which specifically includes: establishing a high-voltage circuit breaker motor drive mechanism fault diagnosis model (second model) capable of describing sensor faults, introducing a sub-vector in the parameter vector of the sensor fault into the high-voltage circuit breaker motor drive mechanism model, and obtaining a sub-model in the high-voltage circuit breaker motor drive mechanism fault diagnosis model of the sensor fault reflecting the fault type corresponding to the current sub-vector; when all sub-vectors in the parameter vector of the sensor fault are introduced into the high-voltage circuit breaker motor drive mechanism model, sub-models of the high-voltage circuit breaker motor drive mechanism fault diagnosis model of all sensor faults are obtained, and all sub-models of the second model together constitute the second model, which is specifically expressed as:

[0035]

[0036] In the formula, f s =[f s,1 f s,2 …f s,r1 ] T The fault parameter vector representing the sensor fault, where

[0037]

[0038] That is, f s,i The value ranges from 0 (normal) to 1 (complete failure).

[0039] The established model (third model) capable of describing the fault model of the system component of the high-voltage circuit breaker motor drive mechanism is introduced into the high-voltage circuit breaker motor drive mechanism model after a sub-vector in the parameter vector of the system component fault is introduced into the high-voltage circuit breaker motor drive mechanism model, and a sub-model in the high-voltage circuit breaker motor drive mechanism system component fault model reflecting the fault type corresponding to the current sub-vector is obtained. When all sub-vectors in the parameter vector of the system component fault are introduced into the high-voltage circuit breaker motor drive mechanism model, sub-models of all high-voltage circuit breaker motor drive mechanism system component fault models are obtained. All sub-models of the third model together constitute the third model, which is specifically expressed as follows:

[0040]

[0041] In the formula, f a =[f a,1 fa,2 …f a,r2 ] T A fault parameter vector representing the extent to which the performance of a system component deviates from its normal value, where

[0042]

[0043] That is, f a,i The value ranges from 0 (normal) to 1 (complete failure).

[0044] Assume that there are r fault diagnosis models for the motor drive mechanism of the high-voltage circuit breaker, and they can be represented by a set {m0 m1…m r} to describe these models:

[0045]

[0046] Where r represents the number of all sub-models. The number and type of fault diagnosis models here can be adjusted according to actual conditions. For example, the present invention establishes a multi-mode adaptive estimation fault model for a high-voltage circuit breaker motor drive mechanism model. Assume that the model set contains three models {m0, m1, m2}, where m0 corresponds to a normal fault-free model; m1 corresponds to fault model ①, i.e., a sensor fault, which is specifically manifested as a constant deviation of the sensor; and m2 corresponds to fault model ②, i.e., a sticking and jamming fault of the moving contact of the high-voltage circuit breaker.

[0047] S3. Input interaction to the second model and the third model to update their states.

[0048] The state estimation value and covariance update are performed on the second model and the third model, and the state estimation value update of the r fault diagnosis models is expressed as:

[0049]

[0050] The covariance update is expressed as,

[0051]

[0052] In the formula, For variable x i The specific meaning of the state estimation at time k, k|k, is: the k before the symbol “|” refers to time k, and the k after the symbol “|” refers to the system state x based on all measurement data and system dynamic information up to time k. i The best estimate of is the updated state estimate, where the superscript “0” indicates the value after the initial update; P j (k|k) is the model m j The optimal covariance matrix at time k is, is the updated covariance matrix; μij (k) represents the model m at time k j The model probability, μ ij The calculation formula for (k) is,

[0053]

[0054] In the formula, π ij is the state transition probability matrix parameter, which represents the state transition probability matrix of the model m. i Transfer to model m j probability.

[0055] S4. Perform time update, measurement value update and probability update on the second model and the third model after the state update, and fuse the updated states to obtain a fourth model reflecting the fault of the motor drive mechanism of the high-voltage circuit breaker.

[0056] Before performing time update, measurement value update and probability update on the second model and the third model after the state update, the first method is used to process the second model and the third model after the state update to obtain the estimated discrete nonlinear system state of each model.

[0057] Among them, in order to discretize the nonlinear system state, there can be many first methods, such as Kalman filter, particle filter, Gaussian filter, Hippert filter, unscented Kalman filter, etc. First of all, Kalman filter is the most commonly used method for fault diagnosis of high-voltage circuit breaker motor drive mechanism at this stage, but the present invention has explained that when the fault occurs and the pre-assumed fault is greatly different, that is, when the residual deviation is large, the Kalman filter cannot obtain a satisfactory estimation result, and is easily affected by nonlinearity and diverges, which is the solution abandoned by the present invention.

[0058] Furthermore, the particle filter has a higher computational complexity than the unscented Kalman filter. The particle filter requires a large number of particles to approximate the posterior probability density, especially in high-dimensional space, which leads to a huge amount of calculation. In addition, in the particle filter, due to the resampling step, the particle diversity may be reduced, thus causing the sample exhaustion problem.

[0059] Compared with the unscented Kalman filter, the Gaussian filter is usually based on a linearized model, and its performance may be significantly degraded for highly nonlinear systems. In addition, the Gaussian filter assumes that the posterior probability distribution is a Gaussian distribution, which may not always hold true in actual systems, especially in the presence of multiple modes or non-Gaussian noise.

[0060] Compared with the unscented Kalman filter, the design of the Hibbert filter is more complicated. The design of the Hibbert filter is usually more complicated than that of the unscented Kalman filter, requiring more prior knowledge and experience. The Hibbert filter needs to solve an optimization problem in each iteration, which may result in a high computational cost. And the Hibbert filter is designed to provide robustness in the presence of model uncertainty and external interference, but this may also result in performance that is not as good as the unscented Kalman filter under the exact model.

[0061] Therefore, the present invention finally considers that the unscented Kalman filter is used for processing and analysis to improve the rapidity and accuracy of fault diagnosis when the fault that occurs is greatly different from the pre-assumed fault, specifically including:

[0062] For model m j (j=1,…,r) to design the unscented Kalman filter. Generate 2n+1 Sigma (discrete nonlinear system state) points for each fault model.

[0063]

[0064] (1) Unscented Kalman Filter Time Update

[0065] Estimates Update:

[0066]

[0067] In the formula, χ j,i (k|k) is the Sigma point χ j,i The optimal value at time k; j,i (k+1|k) is the Sigma point χ j,i The optimal value at time k+1; u(k|k) is the optimal control input at time k; f a (k|k) is the fault parameter vector at time k; W i Represents χ j,i The weight factor of (k+1|k); is the updated model m j The state value of , where the superscript “-” indicates the use of Sigma point χ j,i The updated value.

[0068] Error covariance update:

[0069]

[0070] In the formula, R w (k|k) is the process noise covariance matrix at time k; is the updated covariance matrix.

[0071] (2) Measurement update of unscented Kalman filter

[0072] Generate the corresponding Sigma point Z of each model measurement equation j,i And calculate the forecast estimates:

[0073]

[0074] In the formula, Z j,i (k+1|k) is χ j,i (k+1|k) corresponding predicted estimated value; is the updated model m j The measured value of

[0075] Calculate the Kalman gain:

[0076]

[0077] In the formula, (P zz ) j Z j,i (k+1|k) and The covariance matrix between xz ) j is j,i (k+1|k) and The covariance matrix between j (k+1) is model m j The Kalman gain.

[0078] Estimates Update:

[0079]

[0080] In the formula, r j (k+1|k) is the measured value z j (k+1|k) and its estimated value The residual between is the updated model m j The estimated value of the state.

[0081] Error covariance update:

[0082]

[0083] (3) Probability Update

[0084] Each model is transformed into a likelihood function as follows:

[0085]

[0086] In the formula, δ j Represents model m j The fault object; rj For model m j Corresponds to the residual of the unscented Kalman filter.

[0087] The probability model is updated using the second method, wherein there may be many second methods, such as maximum likelihood estimation, expected utility theory, evidence theory, Bayesian reasoning and the like.

[0088] However, according to our experience in practical applications, maximum likelihood estimation ignores prior information compared to Bayesian inference. Maximum likelihood estimation does not consider any prior knowledge and only relies on observed data. This leads to inaccurate results when data is scarce or biased. And when diagnosing faults in high-voltage circuit breaker motor drive mechanisms, maximum likelihood estimation cannot capture the complexity and diversity of faults because it usually assumes that the data follows a specific distribution.

[0089] Compared with Bayesian reasoning, expected utility theory usually assumes that utility can be linearly added, but in actual decision-making, people may show nonlinear risk attitudes. In addition, in the actual application of fault diagnosis of high-voltage circuit breaker motor drive mechanism, it may be very difficult to determine the utility value of the fault consequence, especially when the consequence of the fault is unclear.

[0090] Compared with Bayesian reasoning, evidence theory lacks commutativity. The combination rules in evidence theory do not satisfy commutativity, which means that the order of evidence will affect the final result. When dealing with complex systems, it may be necessary to define a large number of focal elements (a collection of basic probability distributions), which may lead to computational difficulties.

[0091] Furthermore, when diagnosing the fault of the motor drive mechanism of a high-voltage circuit breaker, the combination rules of the evidence theory may lead to unreasonable results when there is a conflict of evidence, which has a great impact on the fault diagnosis.

[0092] Therefore, Bayesian reasoning is better in this case because it can naturally integrate prior knowledge and new observation data, provide uncertainty quantification, and adapt to the complexity and uncertainty in fault diagnosis.

[0093] The present invention uses Bayesian reasoning to update the probability model expressed as:

[0094]

[0095] (4) State Fusion

[0096] The estimated values ​​and error covariance of each model at each scale are fused, and the fusion calculation formula of the state estimated value is:

[0097]

[0098] Error covariance fusion:

[0099]

[0100] The above formula is the fourth model.

[0101] S5. Introduce an active matching mechanism into the fourth model, correct the model parameters, and predict the fault type based on the corrected model.

[0102] Example 2, reference Figure 3 , which is the second embodiment of the present invention, and is different from the first embodiment in that: a method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism also includes the matching mechanism mentioned in S5 above, which is introduced because the fault diagnosis algorithm can be improved by introducing an active matching mechanism, so as to reduce the number of various fault diagnosis models in the fault model set and improve the accuracy and rapidity of fault diagnosis. The working principle of the introduced active matching mechanism is as follows.

[0103] Each type of fault in the motor drive mechanism of the high-voltage circuit breaker is described by only one fault parameter. The fault parameter assignment does not distinguish between the fault levels of "minor", "more serious", "serious", "particularly serious", etc. In the initial model library, the value of the fault parameter is a random number between 0.05 and 0.95. The computational complexity of the improved fault diagnosis method is Ο(Mn 2 ).

[0104] Substitute the fault parameters into the fourth model and calculate the fault probability of each sub-model of the fourth model. If the probability of failure of this model is greater than the probability of failure of other models and greater than 5%, then it is necessary to adjust the fault parameters in this fault model. If the original value of the fault parameter is too small (close to 0), it is necessary to increase the value of the fault parameter. Otherwise, it is necessary to reduce the value of the fault parameter. Find the most likely failure mode and solve the expression as follows:

[0105]

[0106] like Figure 3 As shown, first keep the fault parameters unchanged and obtain the fault probability value, μ j (k), j∈{1,…,r}; secondly, find the maximum value among them, that is, Then, the fault probability matching is performed. The matching process is specifically divided into three cases. In case 1, if the maximum fault probability is less than a given threshold, such as μ ii (k)<5%, the failure probability remains unchanged, that is, the failure probability increment is 0, that is, Δ(k+1)=0; Case 2, if the maximum failure probability is less than the given threshold and the failure probability at the current moment is greater than the failure probability at the previous moment, such as μ ii (k) ≥ 5% and μ ii(k)≥μ ii (k-1) are satisfied at the same time, the failure probability increases, that is, Δ(k+1)=0.1; Case three, if the maximum failure probability is less than the given threshold and the failure probability at the current moment is less than the failure probability at the previous moment, the failure probability decreases, that is, Δ(k+1)=-0.1; Finally, the failure probability increment is used to correct the failure probability at the next moment.

[0107] Embodiment 3 is the third embodiment of the present invention, which is different from the first two embodiments in that in order to verify and illustrate the technical effects adopted in this method, this embodiment adopts a traditional technical solution and compares it with the method of the present invention, and compares the test results by means of scientific argumentation to verify the real effect of this method.

[0108] Experimental purpose: To compare the performance of a multi-model adaptive estimation fault diagnosis method based on interactive multi-model and unscented Kalman filtering (the method of the present invention) and a traditional multi-model adaptive estimation method (the prior art solution) in diagnosing sudden faults in a high-voltage circuit breaker motor drive mechanism.

[0109] 1. Experimental methods:

[0110] Sample: A certain number of high voltage circuit breaker motor drive mechanisms are selected, some of which are known to have faults.

[0111] Fault simulation: Simulate different fault conditions on the motor drive mechanism, including faults similar to the pre-assumed faults (the characteristics of this type of fault are very close to the pre-set fault model, that is, the manifestation, occurrence mechanism and impact of the fault are consistent with the model assumptions.) and faults with significant differences (the characteristics of this type of fault are very different from the pre-set fault model, and the manifestation, occurrence mechanism or impact of the fault is not completely consistent with the model assumptions, or may even be completely different).

[0112] Data collection: Collect measurement parameters when the fault occurs.

[0113] Diagnosis implementation: The fault is diagnosed using the method of the present invention and the prior art solution respectively.

[0114] Performance evaluation: The diagnostic accuracy, diagnostic time, computational effort, and ability to approximate nonlinear characteristics of each method were recorded.

[0115] 2. Experimental indicators:

[0116] Accuracy: The ratio of the number of correctly diagnosed faults to the total number of faults.

[0117] Diagnostic time: the time from when a fault occurs to when the diagnostic results are output.

[0118] Computational Amount: The computing resources required to complete the diagnosis.

[0119] Nonlinear approximation capability: the degree to which the filter approximates the nonlinear characteristics of the state equation.

[0120] The experimental data obtained are shown in Table 1.

[0121] Table 1 Experimental data comparison table

[0122]

[0123] Accuracy: As can be seen from the table, the accuracy of the method of the present invention on both types of faults is higher than that of the prior art solutions, especially when the fault is significantly different from the pre-assumption, the accuracy of the method of the present invention is significantly improved.

[0124] Diagnosis time: The diagnosis time of the method of the present invention is generally lower than that of the prior art solutions, indicating that it has an advantage in rapidity.

[0125] Calculation amount: The calculation amount required by the method of the present invention is lower than that of the prior art solution, which means that its calculation efficiency is higher.

[0126] Nonlinear approximation capability: The method of the present invention is superior to existing technical solutions in nonlinear approximation capability, which is particularly important for dealing with common nonlinear problems in practice.

[0127] Example 4, reference Figure 4 , which is the fourth embodiment of the present invention, and is different from the previous three embodiments in that: a high-voltage circuit breaker motor drive mechanism fault diagnosis method includes a model building module 100, a state update and fusion module 200 and a correction module 300; the model building module 100 builds a first model reflecting the high-voltage circuit breaker motor drive, introduces a parameter vector of a sensor fault and a parameter vector of a system component fault into the first model respectively, and builds a second model reflecting the sensor fault of the high-voltage circuit breaker motor drive mechanism, and a third model reflecting the system component fault of the high-voltage circuit breaker motor drive mechanism; the state update and fusion module 200 updates the state of the second model and the third model by interactive input, updates the time, measurement value and probability of the second model and the third model after the state update, and fuses the updated state to obtain a fourth model reflecting the high-voltage circuit breaker motor drive mechanism fault; the correction module 300 introduces an active matching mechanism into the fourth model, corrects the model parameters, and predicts the fault type based on the corrected model.

[0128] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0129] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0130] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0131] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0132] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for diagnosing faults in a high-voltage circuit breaker motor drive mechanism, characterized in that: include, constructing a first model reflecting a high voltage circuit breaker motor drive; Introducing the parameter vector of the sensor failure and the parameter vector of the system component failure into the first model respectively, constructing a second model reflecting the sensor failure of the motor drive mechanism of the high-voltage circuit breaker and a third model reflecting the system component failure of the motor drive mechanism of the high-voltage circuit breaker; Input interaction to the second model and the third model to update the state; Performing time updating, measurement value updating and probability updating on the second model and the third model after the state update, and fusing the updated state to obtain a fourth model reflecting the fault of the motor drive mechanism of the high-voltage circuit breaker; An active matching mechanism is introduced into the fourth model, the model parameters are corrected, and the fault type prediction is performed based on the corrected model.

2. A method for diagnosing faults in a high-voltage circuit breaker motor drive mechanism according to claim 1, characterized in that: The second model and the third model include several sub-models, and the number of the sub-models is determined by the parameter vector, specifically including: The parameter vector of the sensor fault and the parameter vector of the system component fault contain multiple sub-vectors. After introducing a sub-vector of the parameter vector of the sensor fault into the first model, a sub-model in the second model reflecting the fault type corresponding to the current sub-vector is obtained. When all sub-vectors of the parameter vector of the sensor fault are introduced into the first model, all sub-models of the second model are obtained, and all sub-models of the second model together constitute the second model. After introducing a sub-vector in the parameter vector of the system component failure into the first model, a sub-model in the third model that reflects the fault type corresponding to the current sub-vector is obtained. When all sub-vectors in the parameter vector of the system component failure are introduced into the first model, all sub-models of the third model are obtained, and all sub-models of the third model together constitute the third model.

3. A method for diagnosing faults in a high-voltage circuit breaker motor drive mechanism as claimed in claim 2, characterized in that: The inputting interaction into the second model and the third model for state updating includes updating the state estimation value and the covariance of the second model and the third model.

4. A method for diagnosing faults of a high-voltage circuit breaker motor drive mechanism as claimed in claim 3, characterized in that: Before the second model and the third model after the state update are updated in time, measurement value and probability, the method further includes: The second model and the third model after the state update are processed by the first method to obtain the estimated discrete nonlinear system state of each model; The time updating includes updating the estimated value and the error covariance on the time scale of each model based on the first method.

5. A method for diagnosing faults of a high-voltage circuit breaker motor drive mechanism as claimed in claim 4, characterized in that: The measurement value updating includes obtaining a predicted estimated value of the measurement equation of each model under the estimated discrete nonlinear system state based on the first method, obtaining a gain according to the predicted estimated value, and updating the estimated value and error covariance of the measurement equation of each model based on the gain; The probability updating includes converting each model into a likelihood function, and updating the likelihood function into a probability model using a second method; The fourth model reflecting the fault of the motor drive mechanism of the high-voltage circuit breaker is obtained by fusing the updated states, including fusing the estimated values ​​and error covariances at each scale of each model, and embedding the probability model to obtain the fourth model reflecting the fault of the motor drive mechanism of the high-voltage circuit breaker. The fourth model has the same structure as the second model and the third model, and includes several fourth model sub-models.

6. A method for diagnosing faults in a high-voltage circuit breaker motor drive mechanism as claimed in claim 5, characterized in that: The active matching mechanism is introduced into the fourth model and the model parameters are corrected, including representing each fault type with a fault parameter, wherein the fault parameter is a numerical range, substituting the fault parameter into the fourth model, calculating the fault probability of each fourth model sub-model, and obtaining the average probability of the fault probability; if the fault probability of any fourth model sub-model is greater than the average probability and higher than a threshold value, the fault parameter in the current fourth model sub-model is corrected so that the fault probability of the current fourth model sub-model meets the threshold requirement.

7. A method for diagnosing faults in a high-voltage circuit breaker motor drive mechanism according to claim 6, characterized in that: The introducing an active matching mechanism into the fourth model and correcting the model parameters also includes: keeping the fault parameters unchanged, calculating the fault probability of each fourth model sub-model at the next moment, and obtaining the average probability at the next moment; if the fault probability of the fourth model sub-model at any next moment is greater than the average probability at the next moment and higher than a threshold, and the fault probability of the fourth model sub-model at the next moment is greater than the fault probability of the fourth model sub-model at the previous moment, then correcting the fault probability of the fourth model sub-model at the current next moment so that the fault probability of the fourth model sub-model at the current next moment meets the threshold requirement.

8. A method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism according to any one of claims 1 to 7, characterized in that: It includes a model building module (100), a state updating and fusion module (200) and a correction module (300); The model building module (100) builds a first model reflecting the high-voltage circuit breaker motor drive, introduces a parameter vector of a sensor fault and a parameter vector of a system component fault into the first model respectively, builds a second model reflecting a sensor fault of the high-voltage circuit breaker motor drive mechanism, and a third model reflecting a system component fault of the high-voltage circuit breaker motor drive mechanism; The state updating and fusion module (200) inputs interactive state updates to the second model and the third model, performs time updates, measurement value updates and probability updates on the second model and the third model after the state updates, and fuses the updated states to obtain a fourth model reflecting the fault of the high-voltage circuit breaker motor drive mechanism; The correction module (300) introduces an active matching mechanism into the fourth model, corrects model parameters, and performs fault type prediction based on the corrected model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a method for diagnosing faults of a high-voltage circuit breaker motor drive mechanism according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a method for diagnosing a fault of a high-voltage circuit breaker motor drive mechanism according to any one of claims 1 to 7 are implemented.

Images