Prediction model training method, prediction method and prediction system for predicting residual mechanical life of circuit breaker, readable storage medium and computer program product

Abstract

The invention relates to a prediction model training method, a prediction method and a prediction system for predicting the residual mechanical life of a circuit breaker, a readable storage medium and a computer program product, and belongs to the technical field of circuit breakers. The method comprises the following steps of: performing weighted average on first mechanical life data and second mechanical life data, and sequencing according to a set sequence to obtain a sample sequence as an original time sequence for establishing a gray model, and performing one-time accumulation on the sample sequence to generate an incremental sequence meeting a quasi-exponential law; solving a grey model established by the progressive increase sequence to obtain a prediction model; or performing one-time cumulative subtraction on the sample sequence to generate a decline sequence meeting the quasi-exponential law, and then solving the grey model established by the decline sequence to obtain the prediction model. According to the method, the first mechanical life data and the second mechanical life data are sorted after weighted averaging, the grey theoretical model is established by using the grey system theory, the grey theoretical model is solved to obtain the prediction model, and the model training is simple.

Description

Technical Field

[0001] The invention relates to a prediction model training method, a prediction method, a prediction system, a readable storage medium and a computer program product for predicting the remaining mechanical life of a circuit breaker, and belongs to the technical field of circuit breakers. Background Art

[0002] According to statistical analysis, the theoretical mechanical life of a circuit breaker device is to successfully complete 10,000 mechanical operations.

[0003] Generally, during the development of new products, type tests are required to verify whether the mechanical life of the circuit breaker equipment meets the theoretical requirements. However, when the product (circuit breaker equipment) leaves the factory, in order to avoid excessive and ineffective consumption of the mechanical life of the circuit breaker equipment, most factory tests are conducted on samples of the same batch of circuit breaker equipment. Therefore, not enough tests are conducted on each circuit breaker equipment to verify whether the mechanical life of the circuit breaker equipment meets the theoretical requirements.

[0004] After the product is put into use in the substation, during the operation and maintenance of the circuit breaker equipment of the substation, the user is not clear whether the mechanical life of the circuit breaker that has been put into use meets the theoretical requirements. It is also difficult to timely and accurately grasp the remaining mechanical life of the circuit breaker equipment during the operation and maintenance process. However, in order to ensure the safe and reliable operation of the circuit breaker equipment, most of the circuit breakers are replaced before the remaining mechanical life of the circuit breaker equipment is exhausted, resulting in high operation and maintenance costs of the circuit breaker equipment and waste of circuit breaker equipment resources.

[0005] The Chinese invention patent with publication number CN115128441A discloses a method for predicting the electrical life and assessing the health status of circuit breakers. The scheme uses the Wiener process and SVR support vector machine model to process data based on the vibration signals and various electrical signals during the opening and closing process of the circuit breaker according to the characteristics of the vibration signals and electrical signals. Finally, a dynamic weighting algorithm is used to assign weights to each model and can be corrected at any time. The circuit breaker can be comprehensively and reliably evaluated and predicted, which improves the accuracy of the circuit breaker electrical life prediction and health status assessment. However, the data processing of this scheme is relatively complex, resulting in relatively high prediction costs.

[0006] The invention patent with publication number WO2024020960Al discloses a method, electronic device and storage medium for predicting the remaining service life of a circuit breaker. The scheme performs N training processes on a neural network model based on parameter values ​​of aging indicators of the circuit breaker collected in at least two time periods to obtain N aging prediction models, which are used to predict the aging state of the circuit breaker; in each training process: the parameter value is input into the neural network model as training data in the training process to train the neural network model into the aging prediction model obtained in the training process; N is a positive integer of at least 2; from the N aging prediction models, an aging prediction model that meets predetermined conditions is selected, and based on the selected aging prediction model, the remaining service life of the circuit breaker is predicted; however, the scheme obtains N prediction models by performing N training processes, and the number of training times of the prediction model is too large. The prediction model that meets the conditions needs to be selected from the N prediction models. The method of obtaining the prediction model that meets the conditions is relatively complicated, resulting in relatively high acquisition costs of the prediction model that meets the conditions for prediction and the prediction cost of predicting the remaining service life of the circuit breaker. Summary of the invention

[0007] The purpose of the present invention is to provide a prediction model training method for predicting the remaining mechanical life of a circuit breaker, so as to solve the problem of the complex method of obtaining the existing prediction model. A prediction method, a prediction system, a readable storage medium and a computer program product for predicting the remaining mechanical life of a circuit breaker are also provided to solve the problem of the complex method of obtaining the prediction model used to predict the remaining mechanical life of a circuit breaker.

[0008] To achieve the above object, the solution of the present invention includes:

[0009] A prediction model training method for predicting the remaining mechanical life of a circuit breaker of the present invention comprises:

[0010] The first mechanical life data of the circuit breaker of the same type under different vibration signals and the second mechanical life data under different sound signals are weighted averaged to obtain sample data, and the sample data are arranged in a set order to obtain a sample series;

[0011] An increasing sequence or a decreasing sequence is generated by performing one accumulation or one accumulation on the sample sequence, a grey model is established for the increasing sequence or the decreasing sequence, and the grey model is solved to obtain a prediction model which takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample sequence as input and the accumulated mechanical life data of the increasing sequence as output, or takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample sequence as input and the accumulated mechanical life data of the decreasing sequence as output.

[0012] Furthermore, the first mechanical life data and the second mechanical life data are both used to characterize the number of mechanical operations currently remaining for the corresponding circuit breaker.

[0013] Furthermore, the first mechanical life data is obtained by subtracting the number of mechanical operations currently completed by the circuit breaker under the corresponding vibration signal from the theoretical mechanical life of the total number of mechanical operations completed by the circuit breaker of this model under different vibration signals; the second mechanical life data is obtained by subtracting the number of mechanical operations currently completed by the circuit breaker under the corresponding sound signal from the theoretical mechanical life of the total number of mechanical operations completed by the circuit breaker of this model under different sound signals.

[0014] Furthermore, the vibration signal is collected by a vibration sensor installed on the outer side wall of the circuit breaker cylinder.

[0015] Furthermore, the sound signal is collected by a sound sensor installed on the outer side wall of the circuit breaker cylinder.

[0016] A method for predicting the remaining mechanical life of a circuit breaker according to the present invention comprises:

[0017] The serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series are input into the prediction model to obtain the cumulative data of the mechanical life in an increasing series or the cumulative data of the mechanical life in a decreasing series. The cumulative data of the mechanical life are restored by a cumulative-subtraction algorithm or the cumulative-subtraction data of the mechanical life are restored by a cumulative-addition algorithm to obtain a predicted value for predicting the remaining mechanical life of the circuit breaker to be tested; the prediction model is trained by the prediction model training method for predicting the remaining mechanical life of the circuit breaker as mentioned above, and the circuit breaker model type used for the first mechanical life data and the second mechanical life data in the prediction model training method is the same as the model type of the circuit breaker to be tested.

[0018] Furthermore, the vibration signal is collected by a vibration sensor installed on the outer wall of the circuit breaker cylinder; and the sound signal is collected by a sound sensor installed on the outer wall of the circuit breaker cylinder.

[0019] A prediction system for predicting the remaining mechanical life of a circuit breaker of the present invention comprises a processor, and the processor is used to execute a computer program to implement the steps of the prediction method for predicting the remaining mechanical life of a circuit breaker as described above.

[0020] A computer-readable storage medium of the present invention stores a computer program, which implements the steps of the above-mentioned method for predicting the remaining mechanical life of a circuit breaker when executed by a processor.

[0021] A computer program product of the present invention comprises a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for predicting the remaining mechanical life of a circuit breaker are implemented.

[0022] Beneficial effects of the present invention:

[0023] The present invention is a pioneering invention, which provides a prediction model training method for predicting the remaining mechanical life of a circuit breaker. The method uses a sample series obtained by weighted averaging the first mechanical life data and the second mechanical life data and sorting them in a set order as the original time data series (original time series) for establishing a grey system theoretical model (grey model). Since the establishment of the grey model requires processing the irregular sample series into a series that satisfies the quasi-exponential law, the sample series is accumulated once to generate an increasing series that satisfies the quasi-exponential law, and then the grey model established by the increasing series is solved to obtain: a prediction model that uses the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series as input and uses the accumulated data of the mechanical life of the increasing series as output;

[0024] Or by performing a cumulative reduction on the sample series to generate a decreasing series that satisfies the quasi-exponential law, and then by solving the grey model established by the decreasing series, a prediction model is obtained: the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series are used as input, and the cumulative reduction data of the mechanical life of the decreasing series is used as output.

[0025] The present invention only needs to sort the data after weighted averaging the first mechanical life data and the second mechanical life data, and then process the irregular sample series into a series that satisfies the quasi-exponential law, so as to establish a grey theoretical model using grey system theory, and then train the prediction model by solving the grey theoretical model. The training method of the prediction model is simple and reliable, and can predict the number of completed mechanical operations or the remaining number of mechanical operations when the circuit breaker generates a certain sound signal and a certain vibration signal, so as to predict the remaining mechanical life of the circuit breaker.

[0026] Also provided is a prediction method, prediction system, readable storage medium and computer program product for predicting the remaining mechanical life of a circuit breaker, wherein the sequence numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample sequence are input into the prediction model to obtain the mechanical life cumulative data of the increasing sequence, and the mechanical life cumulative data is restored by a cumulative-subtraction algorithm to obtain a predicted value;

[0027] Or by inputting the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series into the prediction model, the cumulative reduction data of the mechanical life in a decreasing series is obtained, and the cumulative reduction data of the mechanical life is restored by an accumulation algorithm to obtain the predicted value.

[0028] The predicted value is used to predict the remaining mechanical life of the circuit breaker to be tested. Since the training method of the prediction model is simple and reliable, it can effectively reduce the prediction cost of predicting the remaining mechanical life of the circuit breaker to be tested. Among them, the circuit breaker to be tested and the circuit breaker used for training the prediction model are of the same model.

[0029] Among them, the first mechanical life data and the second mechanical life data are obtained using big data to meet the richness of the weighted average of the first mechanical life data and the second mechanical life data as sample data, which can cover the mechanical life data of the circuit breaker under different vibration signals and different sound signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a line graph of the original series and the accumulated series. DETAILED DESCRIPTION

[0031] In order to solve the problem in the background technology, the present invention provides a prediction model training method for predicting the remaining mechanical life of a circuit breaker. The method obtains a sample series by sorting sample data that characterizes the coupling effect of vibration signals and sound signals on the mechanical life of the circuit breaker. The irregular sample series is processed into a series that satisfies the quasi-exponential law. Then, a gray model is established using the series that satisfies the quasi-exponential law. The gray model is solved to obtain a prediction model for predicting the remaining mechanical life of the circuit breaker, thereby realizing the training of the prediction model.

[0032] A prediction method, prediction system, readable storage medium and computer program product for predicting the remaining mechanical life of a circuit breaker are also provided. The prediction value is obtained by restoring the result output by the above prediction model to realize the prediction of the remaining mechanical life of the circuit breaker.

[0033] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0034] An embodiment of a method for predicting the remaining mechanical life of a circuit breaker:

[0035] A method for predicting the remaining mechanical life of a circuit breaker, comprising:

[0036] The serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker under test during mechanical operation in the sample series are input into the prediction model to obtain the cumulative data of the mechanical life in an increasing series. The cumulative data of the mechanical life is restored using a cumulative-subtraction algorithm to obtain a predicted value for predicting the remaining mechanical life of the circuit breaker under test.

[0037] The prediction model is obtained as follows:

[0038] The sample series is accumulated once to generate an increasing series, a grey model is established for the increasing series, and the grey model is solved to obtain a prediction model which takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series as input and the accumulated data of the mechanical life of the increasing series as output.

[0039] As another embodiment, a method for predicting the remaining mechanical life of a circuit breaker includes:

[0040] The serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker under test during mechanical operation in the sample series are input into the prediction model to obtain the cumulative reduction data of the mechanical life in a decreasing series. The cumulative reduction data of the mechanical life is restored by an accumulation algorithm to obtain a predicted value for predicting the remaining mechanical life of the circuit breaker under test.

[0041] The prediction model is obtained as follows:

[0042] The sample series is subtracted once to generate a decreasing series, a grey model is established for the decreasing series, and the grey model is solved to obtain a prediction model which takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker under test during mechanical operation in the sample series as input and the cumulative reduction data of the mechanical life of the decreasing series as output.

[0043] The sample sequence is obtained as follows:

[0044] The first mechanical life data of the circuit breaker of the same model as the circuit breaker to be tested under different vibration signals and the second mechanical life data under different sound signals are weighted averaged to obtain sample data, and the sample data are arranged in a set order to obtain a sample series.

[0045] Specifically, the first mechanical life data and the second mechanical life data are both used to characterize the number of mechanical operations currently remaining for the corresponding circuit breaker, and the predicted value is used to characterize the number of mechanical operations currently remaining for the circuit breaker to be tested.

[0046] The first mechanical life data is a difference between a theoretical mechanical life used to characterize the total number of mechanical operations completed by the circuit breaker of this model under different vibration signals and the number of mechanical operations currently completed by the circuit breaker under the corresponding vibration signal.

[0047] The second mechanical life data is a difference between the theoretical mechanical life used to characterize the total number of mechanical operations completed by the circuit breaker of this model under different sound signals and the number of mechanical operations currently completed by the circuit breaker under the corresponding sound signal.

[0048] As another implementation, the first mechanical life data and the second mechanical life data are both used to characterize the number of mechanical operations currently completed by the corresponding circuit breaker, and the predicted value is used to characterize the number of mechanical operations currently completed by the circuit breaker to be tested.

[0049] The predicted value predicts the remaining mechanical life of the circuit breaker to be tested in the following way:

[0050] The number of mechanical operations currently remaining for the circuit breaker under test is obtained by obtaining the number of mechanical operations currently completed by the circuit breaker under test. The number of mechanical operations currently remaining for the circuit breaker under test is specifically obtained by subtracting the number of mechanical operations currently completed by the circuit breaker under the corresponding vibration signal and the corresponding sound signal from the theoretical mechanical life representing the total number of mechanical operations completed by the circuit breaker of this model under different vibration signals and different sound signals.

[0051] The first mechanical life data is the number of mechanical operations currently completed by all circuit breakers under the corresponding vibration signal.

[0052] The second mechanical life data is the number of mechanical operations currently completed by all circuit breakers under the corresponding sound signal.

[0053] Specifically, the vibration signal is collected by a vibration sensor installed on the outer side wall of the circuit breaker cylinder.

[0054] Specifically, the sound signal is collected by a sound sensor installed on the outer side wall of the circuit breaker cylinder.

[0055] For ease of understanding, the steps of the prediction method for predicting the remaining mechanical life of a circuit breaker are described with an example:

[0056] A vibration sensor and a sound sensor are installed on the outer side wall of the circuit breaker equipment cylinder. The vibration sensor is used to collect vibration signals during mechanical operation of the circuit breaker equipment; the sound sensor is used to collect sound signals during mechanical operation of the circuit breaker equipment.

[0057] The vibration signals and sound signals generated when all mechanical operations are performed during the circuit breaker equipment test, as well as the vibration signals and sound signals generated when all mechanical operations are performed during the actual operation of the circuit breaker equipment, are collected to obtain the test data on the relationship between the mechanical life of the circuit breaker equipment under different vibration signals and different sound signals and the actual operation data on the relationship between the mechanical life of the circuit breaker equipment under different vibration signals and different sound signals.

[0058] According to any or all of the collected test data and actual operation data of the circuit breaker equipment, the original data sequence of mechanical life of the circuit breaker equipment of the same type under different vibration signals (time domain or frequency domain) (first mechanical life data) is obtained. and the mechanical life raw data sequence (second mechanical life data) of the circuit breaker equipment under different sound signals (loudness, tone or timbre) Among them, the original data of mechanical life Raw data of mechanical life The circuit breaker equipment used is of the same type and model.

[0059] In order to improve the collection efficiency, the vibration signal and the sound signal of the same circuit breaker when performing a certain mechanical operation can also be collected at the same time, that is, t1 = t2.

[0060] Use the weighted average method to give two original data series and Assign weight m to generate a new coupled raw data Among them, the coupled raw data As sample data, Any of the following.

[0061]

[0062] Coupling raw data Sort by the set order to get the sample sequence For ease of understanding, only 9 coupling raw data are used as examples. As shown in Table 1 below:

[0063] Table 1

[0064]

[0065] The coupled raw data The corresponding simulation data are shown in Table 2 below:

[0066] Table 2

[0067]

[0068] The 9 coupled raw data are sorted in the following first order to obtain the sample sequence:

[0069]

[0070] Or sort the 9 coupled raw data in the following second order to obtain the sample sequence:

[0071]

[0072] Or sort the 9 coupled raw data in the third order as follows to obtain the sample sequence:

[0073] Of course, the sorting can also be performed in other ways.

[0074] For the sample series Do an accumulation to generate an increasing sequence Increasing series as follows:

[0075]

[0076] in, Right now:

[0077] Or for the sample series Do a cumulative reduction to generate a decreasing series Decreasing series as follows:

[0078]

[0079] Among them, when k = 1, When k = 2, When k>2,

[0080] Right now:

[0081] like Figure 1 As shown, the sample sequence of the 9 coupled raw data in the above example is sorted in the first order Do an addition to turn the irregular original number series into an increasing number series.

[0082] Let z (1) (t) Yes The mean series of

[0083] z (1) (t)=(z (1) (1),z (1) (2),…,z (1) (k),…z (1) (n))

[0084] in,

[0085]

[0086] In the formula, k = 2, 3,…, n.

[0087] So the GM(1,1) model is:

[0088]

[0089] In the formula, k = 2, 3, ..., n; a is a constant, called the development gray number; b is the endogenous control gray number, which is a constant input to the system.

[0090] Bundle k=2,3,…,n in is regarded as a continuous variable t, and the corresponding whitened differential equation (grey prediction model) is:

[0091]

[0092] remember

[0093] a, b are estimated using the least squares method:

[0094]

[0095] Solve the differential equation, i.e. the prediction model:

[0096]

[0097] When k=1,2,…,n-1, is the fitted value; when k>n-1, is the predicted value, and the sample sequence is obtained by post-subtraction algorithm The fitted and predicted values ​​of .

[0098] View sample series based on vibration and sound signals of mechanical operation of the same type of circuit breaker equipment at any time The corresponding sequence number t (t = 1, 2, 3, ..., n), because the prediction model output by the increasing sequence is a cumulative value at time t, it is necessary to substitute the sequence numbers k1 = t-1 and k2 = t-2 into the prediction model to obtain and Use the post-subtraction algorithm (cumulative subtraction algorithm) to get the sample sequence The predicted value of the circuit breaker equipment at time t for predicting the remaining mechanical life is

[0099] Accordingly, the sample series is viewed based on the vibration signal and sound signal of the same type of circuit breaker equipment during mechanical operation at any time. The corresponding sequence number t (t = 1, 2, 3, ..., n), because the prediction model output by the decreasing sequence is a cumulative decrease value at time t, it is necessary to substitute the sequence numbers k1 = t-1 and k2 = t-2 into the prediction model to obtain and Use the post-addition algorithm (accumulation algorithm) to get the sample sequence The predicted value of the circuit breaker equipment at time t for predicting the remaining mechanical life is

[0100] When both the first mechanical life data and the second mechanical life data adopt mechanical life data representing the currently remaining number of mechanical operations of the corresponding circuit breaker, the obtained prediction value represents the currently remaining number of mechanical operations of the circuit breaker to be tested.

[0101] When the first mechanical life data and the second mechanical life data both use mechanical life data that characterize the number of mechanical operations currently completed by the corresponding circuit breaker, the obtained predicted value characterizes the number of mechanical operations currently completed by the circuit breaker to be tested. The remaining number of mechanical operations of the circuit breaker to be tested is obtained by subtracting the number of mechanical operations currently completed by the circuit breaker under the corresponding vibration signal and the corresponding sound signal from the theoretical mechanical life of the total number of mechanical operations completed by the circuit breaker of this model under different vibration signals and different sound signals. That is, the difference between the ideal mechanical life and the predicted mechanical life of the circuit breaker to be tested is taken as the predicted value of the remaining mechanical life of the circuit breaker to be tested.

[0102] The above-mentioned grey prediction model is also called GM model, which is a model for estimating and predicting the behavior characteristics of a certain indicator related to time within a certain range. Grey prediction mainly uses the GM (1,1) model in grey system theory to predict data.

[0103] Taking into account the test requirements of circuit breaker equipment and the actual needs of users in actual operation and maintenance, a prediction model for predicting the remaining mechanical life of a certain type of circuit breaker equipment is established based on the coupled influence relationship between the two signal characteristics of vibration and sound on the mechanical life of this type of circuit breaker equipment. The remaining mechanical life of circuit breaker equipment can be predicted by using different sound and vibration signal characteristics at any time. The remaining mechanical life of circuit breakers of different voltage levels can be accurately predicted. By using the prediction method of the prediction model, the remaining mechanical life of circuit breaker equipment can be predicted in time, which provides technical reference for circuit breaker equipment test verification or user's remote operation and maintenance of substations, improves remote operation and maintenance efficiency and reduces management costs.

[0104] An embodiment of a prediction model training method for predicting the remaining mechanical life of a circuit breaker:

[0105] A prediction model training method for predicting the remaining mechanical life of a circuit breaker, comprising:

[0106] The sample series is accumulated once to generate an increasing series, a grey model is established for the increasing series, and the grey model is solved to obtain a prediction model which takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series as input and the accumulated data of the mechanical life of the increasing series as output.

[0107] As another implementation method, the sample series is cumulatively subtracted once to generate a decreasing series, a grey model is established for the decreasing series, and the grey model is solved to obtain: a prediction model that takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series as input and takes the cumulative mechanical life data of the decreasing series as output.

[0108] The sample sequence is obtained as follows:

[0109] The first mechanical life data of the circuit breaker of the same model as the circuit breaker to be tested under different vibration signals and the second mechanical life data under different sound signals are weighted averaged to obtain sample data, and the sample data are arranged in a set order to obtain a sample series.

[0110] The above prediction model training method has been described in detail in an embodiment of a prediction method for predicting the remaining mechanical life of a circuit breaker, and will not be described in detail here.

[0111] An embodiment of a prediction system for predicting the remaining mechanical life of a circuit breaker:

[0112] A prediction system for predicting the remaining mechanical life of a circuit breaker includes a processor, and the processor is used to execute a computer program to implement the steps of a prediction method for predicting the remaining mechanical life of a circuit breaker. Among them, a prediction method for predicting the remaining mechanical life of a circuit breaker has been described in detail in an embodiment of a prediction method for predicting the remaining mechanical life of a circuit breaker, and will not be repeated here.

[0113] An embodiment of a computer readable storage medium:

[0114] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a method for predicting the remaining mechanical life of a circuit breaker. A method for predicting the remaining mechanical life of a circuit breaker has been described in detail in an embodiment of a method for predicting the remaining mechanical life of a circuit breaker, and will not be repeated here.

[0115] An embodiment of a computer program product:

[0116] A computer program product includes a computer program, which, when executed by a processor, implements the steps of a method for predicting the remaining mechanical life of a circuit breaker. A method for predicting the remaining mechanical life of a circuit breaker has been described in detail in an embodiment of a method for predicting the remaining mechanical life of a circuit breaker, and will not be repeated here.

Claims

1. A prediction model training method for predicting the remaining mechanical life of a circuit breaker, characterized in that: include: The first mechanical life data of the circuit breaker of the same type under different vibration signals and the second mechanical life data under different sound signals are weighted averaged to obtain sample data, and the sample data are arranged in a set order to obtain a sample series; Perform one accumulation or one accumulation subtraction on the sample sequence to generate an increasing sequence or a decreasing sequence, establish a grey model for the increasing sequence or the decreasing sequence, and solve the grey model to obtain: the prediction model which takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample sequence as input and the accumulated mechanical life data of the increasing sequence as output, or takes the serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample sequence as input and the accumulated mechanical life data of the decreasing sequence as output.

2. The prediction model training method for predicting the remaining mechanical life of a circuit breaker according to claim 1, characterized in that: The first mechanical life data and the second mechanical life data are both used to represent the number of mechanical operations currently remaining for the corresponding circuit breaker.

3. The prediction model training method for predicting the remaining mechanical life of a circuit breaker according to claim 2, characterized in that: The first mechanical life data is obtained by subtracting the number of mechanical operations currently completed by the circuit breaker under the corresponding vibration signal from the theoretical mechanical life of the total number of mechanical operations completed by the circuit breaker of this model under different vibration signals; the second mechanical life data is obtained by subtracting the number of mechanical operations currently completed by the circuit breaker under the corresponding sound signal from the theoretical mechanical life of the total number of mechanical operations completed by the circuit breaker of this model under different sound signals.

4. The prediction model training method for predicting the remaining mechanical life of a circuit breaker according to claim 1, characterized in that: The vibration signal is collected by a vibration sensor installed on the outer side wall of the circuit breaker cylinder.

5. The prediction model training method for predicting the remaining mechanical life of a circuit breaker according to claim 1, characterized in that: The sound signal is collected by a sound sensor installed on the outer side wall of the circuit breaker cylinder.

6. A method for predicting the remaining mechanical life of a circuit breaker, characterized in that: include: The serial numbers corresponding to the vibration signals and sound signals generated by the circuit breaker to be tested during mechanical operation in the sample series are input into the prediction model to obtain the cumulative data of the mechanical life in an increasing series or the cumulative data of the mechanical life in a decreasing series, and the cumulative data of the mechanical life are restored by a cumulative algorithm or the cumulative data of the mechanical life are restored by a cumulative algorithm to obtain a predicted value for predicting the remaining mechanical life of the circuit breaker to be tested; the prediction model is trained by the prediction model training method for predicting the remaining mechanical life of the circuit breaker as described in any one of claims 1 to 3, and the circuit breaker model type used for the first mechanical life data and the second mechanical life data in the prediction model training method is the same as the model type of the circuit breaker to be tested.

7. The method for predicting the remaining mechanical life of a circuit breaker according to claim 6, characterized in that: The vibration signal is collected by a vibration sensor installed on the outer wall of the circuit breaker cylinder; the sound signal is collected by a sound sensor installed on the outer wall of the circuit breaker cylinder.

8. A prediction system for predicting the remaining mechanical life of a circuit breaker, comprising a processor, characterized in that: The processor is used to execute the computer program to implement the steps of the method for predicting the remaining mechanical life of a circuit breaker according to claim 6.

9. 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 the method for predicting the remaining mechanical life of a circuit breaker according to claim 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for predicting the remaining mechanical life of a circuit breaker according to claim 6 are implemented.

Images