Circuit breaker remote monitoring method and device based on multi-source sensor fusion

Abstract

The invention discloses a circuit breaker remote monitoring method and device based on multi-source sensor fusion, and belongs to the field of circuit breakers, and the method comprises the steps: collecting a temperature information array and a humidity information array; collecting a circuit characteristic information sequence, obtaining a predicted circuit characteristic information sequence, and obtaining a predicted temperature information array and a predicted humidity information array; performing overheating aging prediction to obtain predicted mechanical aging information; performing circuit breaker operation fault probability analysis to obtain an open circuit fault probability; and performing failure monitoring analysis on the circuit breaker in combination with the circuit break fault probability based on the prediction accuracy to obtain a circuit break failure probability as a remote monitoring result. The technical problems that in the prior art, remote monitoring precision of the circuit breaker is insufficient, and the failure risk of the circuit breaker cannot be accurately predicted are solved, and the technical effects of accurately monitoring the state of the circuit breaker and accurately evaluating the failure risk are achieved through data fusion and prediction analysis of the multi-source sensor.

Description

Technical Field

[0001] The present invention relates to the field of circuit breakers, and in particular to a circuit breaker remote monitoring method and device based on multi-source sensor fusion. Background Art

[0002] Circuit breakers are important protection and control equipment in power systems, and their reliability is directly related to the safe and stable operation of power systems. Existing remote monitoring technologies for circuit breakers mainly collect the operating parameters of circuit breakers by installing various sensors on circuit breakers, and upload the collected data to the monitoring center for analysis and fault diagnosis. However, the existing technologies have problems such as insufficient monitoring accuracy, inability to accurately predict failure risks, lack of consideration of environmental factors, and low data utilization. Specifically, the existing monitoring methods mainly rely on a single type of sensor data, which makes it difficult to fully reflect the true operating status of the circuit breaker, resulting in insufficient monitoring accuracy and prone to missed reports or false reports. At the same time, the existing fault diagnosis methods are mainly based on threshold judgment or empirical rules, which can only perform post-event diagnosis of faults that have occurred, and cannot predict the future state and failure risk of the circuit breaker, making it difficult to achieve preventive maintenance. In addition, the operating state of the circuit breaker is not only related to its own characteristics, but also affected by environmental factors. The existing technology lacks comprehensive consideration of environmental factors, resulting in insufficient reliability of the diagnosis results. Therefore, the remote monitoring accuracy of circuit breakers in the existing technology is insufficient and the failure risk of circuit breakers cannot be accurately predicted. Summary of the invention

[0003] The present application provides a method and device for remote monitoring of circuit breakers based on multi-source sensor fusion, aiming to solve the technical problems in the prior art of insufficient accuracy of remote monitoring of circuit breakers and inability to accurately predict the risk of circuit breaker failure.

[0004] In view of the above problems, the present application provides a circuit breaker remote monitoring method and device based on multi-source sensor fusion.

[0005] In a first aspect disclosed in the present application, a remote monitoring method for a circuit breaker based on multi-source sensor fusion is provided, the method comprising: based on a temperature sensor array and a humidity sensor array arranged on the circuit breaker, remotely monitoring and collecting temperature information and humidity information of the circuit breaker to obtain a temperature information array and a humidity information array; collecting a circuit characteristic information sequence of a circuit where the circuit breaker is located in the past time, and performing circuit characteristic information prediction to obtain a predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes predicted circuit characteristic information at future time nodes; performing temperature influence analysis on the temperature information array at future time nodes according to the predicted circuit characteristic information sequence to obtain a predicted temperature information array; The system collects the recorded mechanical aging information of the circuit breaker, and performs overheat aging prediction based on the predicted temperature information array to obtain the predicted mechanical aging information; based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information, the system performs the circuit breaker operation failure probability analysis to obtain the circuit breaker failure probability; according to the predicted circuit characteristic information sequence, the circuit breaker protection demand analysis is performed to obtain the circuit breaker demand probability, and combined with the circuit breaker failure probability, based on the prediction accuracy, the circuit breaker failure monitoring analysis is performed to obtain the circuit breaker failure probability as the remote monitoring result.

[0006] Another aspect disclosed in the present application provides a remote monitoring device for a circuit breaker based on multi-source sensor fusion, the device comprising: a multi-source information acquisition unit, for remotely monitoring and acquiring temperature information and humidity information of the circuit breaker based on a temperature sensor array and a humidity sensor array arranged on the circuit breaker, and obtaining a temperature information array and a humidity information array; a circuit feature prediction unit, for acquiring a circuit feature information sequence of the circuit where the circuit breaker is located in the past time, and performing circuit feature information prediction to obtain a predicted circuit feature information sequence, wherein the predicted circuit feature information sequence includes predicted circuit feature information of future time nodes; an environmental impact analysis unit, for performing temperature impact analysis on the temperature information array at future time nodes according to the predicted circuit feature information sequence, and obtaining a predicted temperature information array; The information array is used to analyze the humidity impact of future time nodes based on the predicted temperature information array and the humidity information array to obtain the predicted humidity information array; the mechanical aging prediction unit is used to collect the recorded mechanical aging information of the circuit breaker, and to predict overheat aging based on the predicted temperature information array to obtain the predicted mechanical aging information; the fault probability calculation unit is used to analyze the circuit breaker operation fault probability based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information to obtain the circuit breaker failure probability; the failure risk assessment unit is used to analyze the circuit breaker protection demand according to the predicted circuit characteristic information sequence to obtain the circuit breaker demand probability, and to perform failure monitoring analysis of the circuit breaker based on the prediction accuracy in combination with the circuit breaker failure probability to obtain the circuit breaker failure probability as the remote monitoring result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: By adopting a temperature sensor array and a humidity sensor array arranged on the circuit breaker, the temperature information and humidity information of the circuit breaker are remotely monitored and collected, and a temperature information array and a humidity information array are obtained. By deploying multiple temperature and humidity sensors, a comprehensive monitoring of the temperature and humidity environment of the circuit breaker is achieved, which provides a basis for subsequent data fusion analysis; the circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time is collected, and the circuit characteristic information is predicted to obtain a predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes the predicted circuit characteristic information of future time nodes. By collecting and predicting the historical circuit characteristic information, the circuit breaker is obtained. The characteristic information of the future circuit environment provides a basis for the subsequent temperature and humidity impact analysis; according to the predicted circuit characteristic information sequence, the temperature impact analysis of future time nodes is performed on the temperature information array to obtain the predicted temperature information array; based on the predicted temperature information array and humidity information array, the humidity impact analysis of future time nodes is performed to obtain the predicted humidity information array; using the circuit characteristic prediction information, the future temperature and humidity environment of the circuit breaker is predicted to obtain the predicted temperature and humidity information array, so as to take environmental factors into consideration; the recorded mechanical aging information of the circuit breaker is collected, and overheat aging prediction is performed based on the predicted temperature information array to obtain the predicted mechanical aging information By collecting the historical mechanical aging information of the circuit breaker and combining it with the predicted temperature information, the mechanical aging degree of the circuit breaker is predicted, and the predicted mechanical aging information is obtained, realizing the analysis of the circuit breaker's own characteristics; based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information, the circuit breaker operation failure probability analysis is carried out to obtain the circuit breaker failure probability. Considering the environmental factors and the circuit breaker's own characteristics, the predicted temperature and humidity information and the mechanical aging information are integrated and analyzed to obtain the probability of circuit breaker failure, realizing the accurate evaluation of the circuit breaker status; according to the predicted circuit characteristic information sequence, the circuit breaker protection demand analysis is carried out to obtain the circuit breaker The demand probability is combined with the circuit breaker failure probability, based on the prediction accuracy, and the failure monitoring analysis of the circuit breaker is carried out to obtain the circuit breaker failure probability as the remote monitoring result. On the basis of comprehensively considering the circuit breaker failure probability and the circuit breaker protection demand probability, combined with the prediction accuracy, the circuit breaker failure probability is obtained as the final remote monitoring result, which realizes the technical solution of accurate prediction of the circuit breaker failure risk, solves the technical problems of insufficient accuracy of circuit breaker remote monitoring and inability to accurately predict the circuit breaker failure risk in the existing technology, and achieves the technical effect of accurate monitoring of circuit breaker status and accurate assessment of failure risk through multi-source sensor data fusion and predictive analysis.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flow chart of a circuit breaker remote monitoring method based on multi-source sensor fusion is provided for an embodiment of the present application; Figure 2 A structural schematic diagram of a circuit breaker remote monitoring device based on multi-source sensor fusion is provided for an embodiment of the present application.

[0010] Explanation of the accompanying drawings: multi-source information acquisition unit 11, circuit characteristic prediction unit 12, environmental impact analysis unit 13, mechanical aging prediction unit 14, failure probability calculation unit 15, failure risk assessment unit 16. DETAILED DESCRIPTION

[0011] The overall idea of ​​the technical solution provided by this application is as follows: The embodiments of the present application provide a method and device for remote monitoring of circuit breakers based on multi-source sensor fusion, which collects environmental information of the circuit breaker by deploying a temperature and humidity sensor array, combines the predictive analysis of circuit characteristic information, comprehensively considers the environment in which the circuit breaker is located and its own characteristics, and realizes circuit breaker status evaluation and failure prediction.

[0012] Specifically, first, the temperature and humidity environment of the circuit breaker is fully monitored through the temperature sensor array and the humidity sensor array, and the characteristics of the circuit environment in which the circuit breaker will be located in the future are obtained by predicting the circuit characteristic information. On this basis, through temperature and humidity impact analysis and mechanical aging prediction, the future state of the circuit breaker is accurately predicted and evaluated from the two aspects of environmental factors and its own characteristics. Afterwards, this method deeply integrates and analyzes multi-source data such as the predicted temperature and humidity information, mechanical aging information, and circuit breaker protection requirements to obtain the fault probability and failure probability of the circuit breaker, and realizes accurate monitoring of the circuit breaker state and accurate prediction of failure risks.

[0013] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0014] Embodiment 1, as Figure 1 As shown, an embodiment of the present application provides a circuit breaker remote monitoring method based on multi-source sensor fusion, the method comprising: S1: Based on the temperature sensor array and humidity sensor array arranged on the circuit breaker, remotely monitor and collect the temperature information and humidity information of the circuit breaker to obtain the temperature information array and the humidity information array.

[0015] Specifically, a temperature sensor array and a humidity sensor array are arranged on the circuit breaker. The temperature sensor array is composed of a plurality of temperature sensors, which are distributed at different positions of the circuit breaker, and collects temperature data of various parts of the circuit breaker in real time. Similarly, the humidity sensor array is composed of a plurality of humidity sensors, which are distributed at different positions around the circuit breaker, and collects humidity data of the environment in which the circuit breaker is located in real time. During remote monitoring, the temperature sensor array and the humidity sensor array arranged on the circuit breaker are connected through remote communication methods (such as wireless communication, wired communication, etc.), and the temperature and humidity information collected by the temperature sensor array and the humidity sensor array are remotely obtained, and the temperature information array reflecting the current temperature distribution state of the circuit breaker and the humidity information array reflecting the current humidity distribution state of the environment in which the circuit breaker is located are obtained in real time, and the real-time working status data of the circuit breaker is intuitively presented.

[0016] S2: Collecting a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time, and predicting the circuit characteristic information to obtain a predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes predicted circuit characteristic information at a future time node.

[0017] Specifically, in order to fully grasp the operating status of the circuit where the circuit breaker is located, the circuit characteristic information sequence of the circuit where the circuit breaker is located in the past period of time is collected. Among them, the circuit characteristic information may include but is not limited to the voltage, current, power and other parameters of the circuit. By collecting the circuit characteristic information within a certain time span, a time series data sequence reflecting the operating characteristics of the circuit is formed. After obtaining the circuit characteristic information sequence, the future circuit characteristic information is predicted. By analyzing the operating characteristics of the circuit where the circuit breaker is located in the past period of time, the circuit characteristic information of the circuit in the future period of time is inferred by using the prediction model to form a predicted circuit characteristic information sequence. The predicted circuit characteristic information sequence contains the predicted circuit characteristic information of several future time nodes (for example, the next 10 minutes, 1 hour, etc.), which provides a basis for realizing forward-looking monitoring and early warning of circuit breakers.

[0018] S3: According to the predicted circuit characteristic information sequence, the temperature information array is subjected to temperature impact analysis at future time nodes to obtain a predicted temperature information array; based on the predicted temperature information array and the humidity information array, the humidity impact analysis at future time nodes is performed to obtain a predicted humidity information array.

[0019] Specifically, after obtaining the predicted circuit characteristic information sequence that reflects the future circuit characteristics, the information sequence is further used to predict and analyze the temperature and humidity of the circuit breaker. This is because changes in characteristic parameters such as circuit voltage and current will have a direct impact on the temperature and humidity of the circuit breaker. By analyzing the correlation between circuit characteristic information and temperature and humidity, the temperature and humidity risks faced by the circuit breaker can be predicted.

[0020] First, the predicted circuit characteristic information sequence is used to analyze the temperature impact of the obtained temperature information array at future time nodes to obtain a predicted temperature information array. The predicted temperature information array reflects the temperature distribution of various parts of the circuit breaker at future time nodes under the influence of the predicted circuit characteristic information, so as to predict whether the circuit breaker may face the risk of local overheating. Then, based on the obtained predicted temperature information array, the impact of temperature changes on the humidity of the environment where the circuit breaker is located is further analyzed. By comprehensively considering the predicted temperature information array and the obtained humidity information array, the humidity changes in the environment around the circuit breaker are predicted and analyzed to obtain a predicted humidity information array. The predicted humidity information array reflects the humidity distribution of the environment where the circuit breaker is located at future time nodes under the dual influence of the predicted circuit characteristic information and temperature changes.

[0021] Through predictive analysis of temperature and humidity, forward-looking monitoring of the internal and external environment of the circuit breaker is achieved, which intuitively reflects the thermal and wet risks faced by the circuit breaker, which is crucial for judging the insulation performance and working reliability of the circuit breaker. It also provides an important basis for subsequent analysis of the mechanical aging degree of the circuit breaker.

[0022] S4: collecting recorded mechanical aging information of the circuit breaker, performing overheat aging prediction based on the predicted temperature information array, and obtaining predicted mechanical aging information.

[0023] Specifically, the mechanical parts of the circuit breaker will age during long-term operation, and temperature is an important factor affecting mechanical aging. By analyzing the relationship between historical mechanical aging data and temperature, a model of the impact of temperature on mechanical aging is established, and the model is used to predict the degree of mechanical aging of the circuit breaker.

[0024] First, the recorded mechanical aging information of the circuit breaker is obtained as the basis for predictive analysis. Specifically, the mechanical aging degree of the circuit breaker during the last mechanical aging prediction is used as a reference, and it is used as the recorded mechanical aging information to reflect the wear, fatigue and other aging conditions that have occurred in the mechanical parts of the circuit breaker. Then, the obtained predicted temperature information array is used to predict the mechanical aging degree of the circuit breaker. The predicted temperature information array reflects the temperature distribution that may be reached in various parts of the circuit breaker in the future, and these temperature changes will have different degrees of aging effects on the mechanical parts. By analyzing the predicted temperature change characteristics and combining factors such as the material properties and structural characteristics of the mechanical parts of the circuit breaker, the degree of influence of temperature on mechanical aging can be determined. After the above analysis, the predicted mechanical aging information is obtained, which comprehensively considers the existing aging degree of the circuit breaker and the temperature aging effects that may be suffered in the future, so as to predict the aging risk of the mechanical parts of the circuit breaker in the future.

[0025] S5: Based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information, a circuit breaker operation failure probability analysis is performed to obtain a circuit breaker failure probability.

[0026] Specifically, the reliability of the circuit breaker depends not only on the working status of each part of the circuit breaker, but also on environmental factors and its own aging degree. In order to comprehensively evaluate the failure risk of the circuit breaker, the three key factors of temperature, humidity and mechanical aging are comprehensively considered, and the prediction information of these factors is integrated and analyzed to obtain a quantitative estimate of the overall failure probability of the circuit breaker.

[0027] First, the predicted temperature information array and the predicted humidity information array, as well as the predicted mechanical aging information, are used as input to comprehensively evaluate the working reliability of the circuit breaker. Among them, the predicted temperature information array and the predicted humidity information array reflect the insulation failure risk that each part of the circuit breaker may face in the future, and the predicted mechanical aging information reflects the wear, deformation and other degradation of the mechanical parts of the circuit breaker. After obtaining the three types of prediction information, the circuit breaker fault analysis model trained based on a large amount of historical fault data is used to estimate the overall failure probability of the circuit breaker. The model analyzes the failure of the circuit breaker under different temperature, humidity and mechanical aging conditions, and establishes the corresponding relationship between these factors and the failure rate of the circuit breaker. When applying the model, the predicted temperature, humidity and mechanical aging information are input into the model to obtain the estimated probability of the circuit breaker failure in the future, that is, the circuit breaker failure probability. The circuit breaker failure probability intuitively characterizes the overall reliability level of the circuit breaker, comprehensively reflects the risk of insulation failure, action failure and other failures of the circuit breaker under the predicted working conditions, and provides an important basis for the state assessment, risk warning and maintenance decision of the circuit breaker.

[0028] S6: Perform circuit breaking protection demand analysis according to the predicted circuit characteristic information sequence to obtain the circuit breaking demand probability, and perform failure monitoring analysis of the circuit breaker based on the prediction accuracy in combination with the circuit breaking failure probability to obtain the circuit breaking failure probability as the remote monitoring result.

[0029] Specifically, the failure risk of a circuit breaker depends not only on its own failure probability, but also on the circuit breaking protection requirements of the circuit in which it is located. When an abnormal situation such as overload or short circuit occurs in the circuit, the circuit breaker needs to cut off the circuit in time to prevent the accident from expanding and equipment damage. However, if the circuit breaker itself fails and cannot operate normally at this time, it may lead to serious failure consequences. Therefore, based on the probability of circuit breaking failure, the circuit breaking protection requirements are further considered. By comprehensively evaluating the two risk factors, the circuit breaking failure probability of the circuit breaker is obtained.

[0030] First, the probability of circuit breaking protection demand in the future is obtained, that is, the probability of circuit breaking demand. The probability of circuit breaking demand reflects the possibility of abnormal conditions such as overload and short circuit that require the circuit breaker to cut off the circuit within the predicted time range. In order to obtain the probability of circuit breaking demand, it is necessary to analyze the correlation between the predicted circuit feature information sequence and the historical circuit breaking protection events. For example, by extracting a large number of operation records of similar circuits and marking the time nodes where the circuit breaking protection action occurs, a sample data set reflecting the corresponding relationship between circuit characteristics and circuit breaking demand can be constructed. The circuit breaking protection demand analysis model is trained with the sample data set to realize the function of estimating the probability of circuit breaking demand according to the predicted circuit feature information sequence. After obtaining the probability of circuit breaking demand, it is comprehensively analyzed with the calculated probability of circuit breaking failure to obtain the estimated value of the probability of circuit breaker failure. The probability of circuit breaker failure reflects the joint risk of the circuit breaking protection demand and the failure of the circuit breaker itself to operate normally within the predicted time range. By analyzing historical data, it is found that circuit breaker failure events are often the result of the combined effect of circuit breaking protection demand and circuit breaker failure. Therefore, the probability of circuit breaker failure is calculated by multiplying the probability of circuit breaker demand and the probability of circuit breaker failure, as the probability of two types of risk events occurring at the same time. Afterwards, the obtained probability of circuit breaker failure is used as the remote monitoring result for technicians to make judgments, early warnings, maintenance, etc.

[0031] Furthermore, the embodiment of the present application also includes: Collecting a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time; According to the historical operation data of the circuit in which the circuit breaker is located, a set of sample historical circuit feature information sequences before multiple historical time points is collected, and a set of sample predicted circuit feature information sequences after multiple historical time points is collected; Using the sample historical circuit feature information sequence set and the sample predicted circuit feature information sequence set to train a circuit feature information predictor; The circuit characteristic information predictor is used to perform circuit characteristic information prediction on the circuit characteristic information sequence to obtain predicted circuit characteristic information sequences at multiple time nodes within the future time node.

[0032] In a feasible implementation, by analyzing the historical operation data of the circuit, a circuit characteristic information predictor is trained, and the circuit characteristic information predictor is used to predict the future characteristic information of the circuit, thereby providing important input data for subsequent analysis of the working status of the circuit breaker.

[0033] First, the circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time is collected, including multiple parameters such as the circuit voltage, current, power, etc. By setting sensors and data acquisition devices on the circuit, the changes of these parameters over a period of time are recorded in real time to form a circuit characteristic information sequence reflecting the past operation status of the circuit. Then, in the historical operation data of the circuit where the circuit breaker is located, multiple specific historical time points are selected, and these time points are used as boundaries to extract the circuit characteristic information sequence samples in the period before them as the sample historical circuit characteristic information sequence set; at the same time, the circuit characteristic information sequence samples in the period after these time points are extracted as the corresponding sample predicted circuit characteristic information sequence set, thereby obtaining a sample data set containing multiple sets of historical information and future information pairs for training the circuit characteristic information predictor.

[0034] Next, a circuit feature information predictor is constructed using the obtained sample historical circuit feature information sequence set and sample predicted circuit feature information sequence set. The circuit feature information predictor takes the historical circuit feature information sequence as input and outputs a mapping model that corresponds to the circuit feature information prediction sequence in the future time period. By selecting a machine learning algorithm, such as a related algorithm for time series prediction, and using a sample data set to train and tune the model, a circuit feature information predictor with high accuracy is obtained. Subsequently, the circuit feature information sequence reflecting the operating state of the circuit over a period of time in the past is input into the circuit feature information predictor, which automatically implements the calculation processes such as feature extraction and timing association, and outputs a predicted circuit feature information sequence that represents the feature information of the circuit in the future, giving the expected change trend of parameters such as voltage and current of the circuit at multiple time nodes in the future.

[0035] The circuit characteristic information predictor is trained through machine learning, so that according to the circuit's operating status in the past, its characteristic parameter changes in the future can be accurately predicted, providing data support for subsequent analysis of circuit failure risks and other applications.

[0036] Furthermore, the embodiment of the present application also includes: Training a sensor information predictor based on historical sensor monitoring data of the circuit breaker, wherein the sensor information predictor includes a temperature information prediction branch and a humidity information prediction branch; The sensor information predictor is used to predict the temperature impact of future time nodes on the temperature information array and the prediction circuit characteristic information sequence to obtain a predicted temperature information array, and the humidity impact of future time nodes on the humidity information array and the prediction circuit characteristic information sequence is predicted to obtain a predicted humidity information array.

[0037] In a feasible implementation, first, a sensor information predictor is constructed based on the historical sensor monitoring data collected during the historical operation of the circuit breaker. The sensor information predictor consists of two parallel prediction branches, including a temperature information prediction branch and a humidity information prediction branch, which are responsible for the prediction tasks of temperature information and humidity information respectively. The temperature information prediction branch takes the historical temperature monitoring data of various parts of the circuit breaker as input, and outputs the temperature prediction value corresponding to the future time node; the humidity information prediction branch takes the historical humidity monitoring data of the environment around the circuit breaker as input, and outputs the humidity prediction value corresponding to the future time node. When training the sensor information predictor, not only the historical data of the temperature and humidity sensors are used, but also the characteristic information such as voltage and current that reflect the operating status of the circuit is integrated. Through the correlation analysis of multi-source heterogeneous data, the intrinsic connection between temperature and humidity changes and circuit operating conditions is explored, thereby improving the accuracy of the prediction.

[0038] Subsequently, the temperature information array collected in real time from each part of the circuit breaker and the predicted circuit characteristic information sequence for a period of time in the future are input into the temperature information prediction branch, which automatically extracts the correlation characteristics between temperature changes and circuit states, and generates a predicted temperature information array that characterizes the temperature distribution of multiple future time nodes. Similarly, the humidity information array collected in real time from the environment around the circuit breaker and the predicted circuit characteristic information sequence are input into the humidity information prediction branch, which comprehensively analyzes the humidity change trend and circuit operating conditions, and outputs a predicted humidity information array that reflects the humidity level for a period of time in the future.

[0039] By utilizing the massive amount of sensor monitoring data from the historical operation of the circuit breaker, the complex nonlinear relationship between state quantities such as temperature and humidity and circuit conditions can be adaptively learned in a data-driven modeling manner. When predicting future states, the influence of circuit operation characteristics can be comprehensively considered, and the prediction has stronger predictive adaptability and robustness. At the same time, the sensor information predictor adopts a parallel dual-branch structure, which can independently perform targeted modeling on the two key state quantities of temperature and humidity, which not only reduces the complexity of a single prediction task, but also captures the interactive influence between the two state quantities, improving the prediction accuracy while taking into account the computational efficiency.

[0040] Furthermore, the embodiment of the present application also includes: According to the historical sensor monitoring data of the circuit breaker, a sample circuit characteristic information sequence set is collected, and a sample basic temperature information array set and a sample basic humidity information array set at the starting time point are collected, and a sample predicted temperature information array set and a sample predicted humidity information array set at the predicted time node are collected; Using the sample circuit characteristic information sequence set, the sample basic temperature information array set, and the sample predicted temperature information array set to train a temperature information prediction branch; The sample circuit characteristic information sequence set, the sample basic humidity information array set and the sample predicted humidity information array set are used to predict the humidity information prediction branch; The temperature information prediction branch and the humidity information prediction branch are combined to obtain a sensor information predictor.

[0041] In a preferred embodiment, first, a set of sample circuit characteristic information sequences reflecting the historical operating state of the circuit in which the circuit breaker is located is extracted, and the set includes time series of circuit voltage, current and other parameters in multiple historical time periods. Then, the time nodes corresponding to the circuit characteristic information sequence are selected, and the temperature monitoring data of each part of the circuit breaker at these time nodes are extracted to form a sample basic temperature information array set; at the same time, the humidity monitoring data of the environment around the circuit breaker at these time nodes are extracted to form a sample basic humidity information array set. In addition, the temperature and humidity monitoring values ​​of the future prediction time nodes corresponding to the above sample data are collected as the target output of the training prediction model, and a sample prediction temperature information array set and a sample prediction humidity information array set are formed respectively. In the training process, the sample circuit characteristic information sequence and the sample basic temperature information array of the corresponding time node are used as input, and the sample prediction temperature information array of the future prediction time node is used as the expected output. The model parameters of the temperature information prediction branch are optimized by supervised learning, so that it can accurately learn the correlation mapping between the circuit state and the temperature change. Similar to the temperature information prediction branch, the sample circuit feature information sequence and the sample basic humidity information array are used as input, and the sample predicted humidity information array is used as the expected output. The model parameters of the humidity information prediction branch are optimized through supervised learning, so that it can accurately capture the intrinsic connection between the circuit operation state and the humidity change law. After that, the temperature information prediction branch and the humidity information prediction branch that have been independently trained are combined to form a complete sensor information predictor. In the sensor information predictor, the two prediction branches work in parallel, responsible for the prediction tasks of temperature and humidity respectively. When the circuit feature information sequence, temperature information array and humidity information array of the circuit breaker at the current moment are input, the predictor can comprehensively consider the influence of various types of information and give the prediction results of temperature and humidity in the future.

[0042] Furthermore, the embodiment of the present application also includes: Collecting the mechanical aging degree of the circuit breaker during the last mechanical aging prediction as the recorded mechanical aging information; According to the mechanical aging test data of the circuit breaker, a sample temperature information array set and a sample mechanical aging degree set are collected to train a mechanical overheat aging analyzer; Based on the mechanical overheat aging analyzer, performing aging analysis on the recorded mechanical aging information to obtain an overheat aging degree; The predicted mechanical aging degree is calculated based on the recorded mechanical aging information and the overheat aging degree as the predicted mechanical aging information.

[0043] In a feasible implementation, by collecting mechanical aging information and temperature monitoring data during the historical operation of the circuit breaker and using these data to train a mechanical overheat aging analyzer, the current mechanical aging state of the circuit breaker can be evaluated and future mechanical aging trends can be predicted.

[0044] First, the mechanical aging degree of the circuit breaker during the last mechanical aging prediction is collected as the initial state of this prediction, that is, the mechanical aging information is recorded. The mechanical aging degree is a comprehensive indicator to measure the degree of degradation of the mechanical components of the circuit breaker. It is calculated by analyzing the action characteristic parameters of the circuit breaker (such as opening and closing time, bounce time, etc.), the vibration and noise of the circuit breaker body and other monitoring data, combined with the mechanical aging model. Then, by conducting an accelerated aging test on the circuit breaker, the change data of its mechanical characteristic parameters under different temperature conditions are collected to form a sample temperature information array set; at the same time, according to the collected mechanical characteristic parameters, the corresponding mechanical aging degree is calculated to form a sample mechanical aging degree set. Using the obtained sample temperature information array set and sample mechanical aging degree set, the mechanical overheat aging analyzer is trained so that it can learn the intrinsic relationship between temperature stress and mechanical aging.

[0045] Next, the collected recorded mechanical aging information is input into the mechanical overheat aging analyzer, which comprehensively considers the existing mechanical aging degree of the circuit breaker and calculates the increment of mechanical aging degree after continuing to operate for a period of time under the current temperature stress, that is, the overheat aging degree. The overheat aging degree reflects the accelerating effect of the high temperature environment on the degradation of the mechanical components of the circuit breaker in the future. Afterwards, the predicted mechanical aging degree is calculated based on the recorded mechanical aging information and the overheat aging degree as the prediction result of the mechanical aging state of the circuit breaker. Among them. The predicted mechanical aging degree is based on the current mechanical aging degree of the circuit breaker, and the influence of overheat aging in the future period is superimposed to comprehensively evaluate the degradation degree of the mechanical components of the circuit breaker at future time nodes. The predicted mechanical aging degree is obtained by weighted summing the recorded mechanical aging information and the overheat aging degree.

[0046] Through data-driven modeling, the nonlinear relationship between high temperature stress and mechanical aging is adaptively learned, and the dynamic impact of temperature changes can be considered when evaluating overheat aging, which has stronger predictive adaptability. At the same time, by collecting mechanical aging information during the historical operation of the circuit breaker and using it as the starting point for prediction, the initial health state of the circuit breaker's mechanical components is accurately characterized, so that the prediction results can objectively reflect the cumulative degradation effect of the circuit breaker and improve the continuity and consistency of the judgment of the mechanical aging trend.

[0047] Furthermore, the embodiment of the present application also includes: According to the predicted temperature information array and the predicted humidity information array, average predicted temperature information and average predicted humidity information are calculated and obtained; According to the circuit breaker fault record data under different parameters, the sample average temperature information set, the sample average humidity information set, and the sample mechanical aging information set are collected, and according to the failure ratio of the circuit breaker under different parameters, the sample failure probability set is marked; Using the sample average temperature information set, the sample average humidity information set, the sample mechanical aging information set and the sample failure probability set to train a circuit breaker failure analyzer; Based on the circuit breaker fault analyzer, the circuit breaker failure probability is predicted based on the average predicted temperature information, the average predicted humidity information and the predicted mechanical aging information to obtain a circuit breaker failure probability.

[0048] In a feasible implementation, multiple influencing factors such as temperature, humidity and mechanical aging are comprehensively considered, and the historical fault data of the circuit breaker is analyzed and modeled to train a circuit breaker fault analyzer, and the circuit breaker fault analyzer is used to predict and evaluate the failure risk of the circuit breaker in the future.

[0049] First, the average predicted temperature information and average predicted humidity information are calculated based on the obtained predicted temperature information array and predicted humidity information array. Since the temperature and humidity monitoring data are presented in the form of an array, reflecting the temperature and humidity distribution of different locations inside the circuit breaker and the external environment, in order to facilitate subsequent fault analysis, these local monitoring data are integrated. By taking the average value of each element in the predicted temperature information array, a representative average predicted temperature information is obtained; similarly, the average predicted humidity information is obtained by taking the average value of each element in the predicted humidity information array. The average predicted temperature information and the average predicted humidity information respectively reflect the temperature and humidity levels that the circuit breaker may be in in the future.

[0050] Then, by statistically analyzing the fault record data of circuit breakers of different models and different operating conditions, the average temperature, average humidity and mechanical aging information in the period before the fault moment are extracted to form a sample average temperature information set, a sample average humidity information set and a sample mechanical aging information set. At the same time, according to the proportion of circuit breaker failures under different parameter combinations, the sample data are labeled to obtain the corresponding sample failure probability set, thereby establishing a sample data set that reflects the association between the circuit breaker failure characteristics and environmental factors and its own health status. Subsequently, during the training process, the average temperature, average humidity and mechanical aging information in the sample data set are used as input features, and the corresponding sample failure probability is used as the expected output, and the model parameters of the fault analyzer are optimized using the supervised learning method. Through training, the circuit breaker fault analyzer can adaptively learn the complex nonlinear relationship between factors such as temperature, humidity and mechanical aging and the risk of circuit breaker failure, and has the ability to predict the failure probability based on environmental conditions and equipment health status.

[0051] Next, the calculated average predicted temperature information and average predicted humidity information, as well as the predicted mechanical aging information, are input into the circuit breaker fault analyzer. The circuit breaker fault analyzer comprehensively considers these influencing factors and calculates the probability of failure of the circuit breaker within the predicted time range, that is, the circuit breaker failure probability, and quantitatively evaluates the overall failure risk of the circuit breaker under specific working conditions in the future.

[0052] Furthermore, the embodiment of the present application also includes: According to the operation data records of the same type of circuits, a set of characteristic information sequences of the same type of circuits is collected, and according to the probability of the occurrence of circuit breaking protection requirements under different characteristic information sequences of the same type of circuits, a set of sample circuit breaking requirements probability is marked; Using the same family circuit feature information sequence set and the sample circuit breaking demand probability set, training a circuit breaking protection demand analyzer, performing circuit breaking protection demand analysis on the predicted circuit feature information sequence, and obtaining a circuit breaking demand probability; Testing the circuit breaker fault analyzer and the circuit breaker protection requirement analyzer to obtain the circuit breaker fault analysis accuracy and the circuit breaker requirement analysis accuracy; According to the circuit breaking demand probability and the circuit breaking failure probability, a basic circuit breaking failure probability is calculated; Based on the circuit breaking fault analysis accuracy and the circuit breaking demand analysis accuracy, a compensation calculation is performed on the basic circuit breaking failure probability to obtain a circuit breaking failure probability as a remote monitoring result.

[0053] In a preferred embodiment, based on the analysis of the failure risk of the circuit breaker itself, the protection requirements of the circuit in which the circuit breaker is located are further considered. By evaluating the joint probability of failure of the circuit breaker when a circuit breaking protection requirement occurs, a circuit breaking failure probability that fully reflects the risk status of the circuit breaker is obtained.

[0054] First, by collecting and counting characteristic parameters such as voltage and current of multiple circuits of the same family, a set of characteristic information sequences of circuits of the same family that reflects their typical operating modes is formed. At the same time, the historical accident records of these circuits of the same family are analyzed, the proportion of circuit-breaking protection actions under different operating conditions is counted, and the sample data is labeled accordingly to obtain a set of sample circuit-breaking demand probabilities that characterize the probability of occurrence of circuit-breaking protection requirements. Then, the circuit-breaking protection requirement analyzer is trained using the constructed set of characteristic information sequences of circuits of the same family and the set of sample circuit-breaking demand probabilities. The circuit-breaking protection requirement analyzer is an intelligent model that takes the circuit characteristic information sequence as input and outputs the probability of circuit-breaking protection requirements under the operating condition. The model is trained and optimized using sample data so that it can adaptively learn the corresponding relationship between circuit operating characteristics and circuit-breaking protection requirements. The obtained predicted circuit characteristic information sequence is input into the trained circuit-breaking protection requirement analyzer to predict the probability of circuit-breaking protection requirements in the circuit in the future, that is, the probability of circuit-breaking requirements.

[0055] In order to evaluate the prediction reliability of circuit breaker fault analysis and circuit breaker protection demand analysis, the corresponding intelligent models were tested. A set of independent historical data samples were used to evaluate the prediction performance of the circuit breaker fault analyzer and the circuit breaker protection demand analyzer. By comparing the degree of consistency between the prediction results and the actual situation, the prediction accuracy of the two types of analyzers, namely the circuit breaker fault analysis accuracy and the circuit breaker demand analysis accuracy, was obtained, reflecting the confidence level of fault analysis and protection demand analysis. Then, based on the circuit breaker failure probability and the circuit breaker demand probability, the joint probability of the two conditions of circuit breaker protection demand and circuit breaker failure occurring simultaneously within the prediction time range was obtained through product calculation, namely the basic circuit breaker failure probability, and the risk level of circuit breaker failure in actual working conditions was evaluated. Considering that both circuit breaker fault analysis and circuit breaker protection demand analysis have certain prediction errors, in order to correct the prediction results of circuit breaker failure probability, compensation calculation was performed based on the obtained basic circuit breaker failure probability and the prediction accuracy of the two types of analyzers. Specifically, the basic circuit breaker failure probability is divided by the circuit breaker fault analysis accuracy and the circuit breaker demand analysis accuracy, and the risk assessment bias caused by the prediction error is offset by amplifying the prediction results. The compensated circuit breaker failure probability is used as the remote monitoring result, reflecting the possibility of circuit breaker failure in the future, making the monitoring result more robust and reliable.

[0056] In summary, the circuit breaker remote monitoring method based on multi-source sensor fusion provided in the embodiment of the present application has the following technical effects: Based on the temperature sensor array and humidity sensor array arranged on the circuit breaker, the temperature information and humidity information of the circuit breaker are remotely monitored and collected to obtain the temperature information array and humidity information array, realize comprehensive monitoring of the temperature and humidity environment of the circuit breaker, and obtain the temperature information array and humidity information array reflecting the real-time working environment of the circuit breaker, which provide basic input for subsequent data analysis and status evaluation. The circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time is collected, and the circuit characteristic information is predicted to obtain the predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes the predicted circuit characteristic information of the future time node, realizes the prediction of the future working environment of the circuit breaker, and provides a key reference basis for the temperature and humidity impact analysis. According to the predicted circuit characteristic information sequence, the temperature impact analysis of the future time node is performed on the temperature information array to obtain the predicted temperature information array. Based on the predicted temperature information array and humidity information array, the humidity impact analysis of the future time node is performed to obtain the predicted humidity information array, which fully considers the impact of the circuit environment on the temperature and humidity of the circuit breaker, and realizes the accurate prediction of the future temperature and humidity environment of the circuit breaker. Collect the recorded mechanical aging information of the circuit breaker, and perform overheat aging prediction based on the predicted temperature information array to obtain the predicted mechanical aging information, realize the predicted evaluation of the mechanical characteristics of the circuit breaker itself, and provide an important basis for the subsequent failure probability analysis. Based on the predicted temperature information array, predicted humidity information array and predicted mechanical aging information, perform the circuit breaker operation failure probability analysis to obtain the circuit breaker failure probability, realize the evaluation of the comprehensive state of the circuit breaker, and quantify the possibility of circuit breaker failure. Perform circuit breaker protection demand analysis based on the predicted circuit characteristic information sequence to obtain the circuit breaker demand probability. Combined with the circuit breaker failure probability, based on the prediction accuracy, perform circuit breaker failure monitoring analysis to obtain the circuit breaker failure probability as the remote monitoring result, and realize the accurate prediction of circuit breaker failure risk.

[0057] Embodiment 2 is based on the same inventive concept as the circuit breaker remote monitoring method based on multi-source sensor fusion in the above embodiment. Figure 2 As shown, an embodiment of the present application provides a circuit breaker remote monitoring device based on multi-source sensor fusion, the device comprising: The multi-source information acquisition unit 11 is used to remotely monitor and acquire the temperature information and humidity information of the circuit breaker based on the temperature sensor array and humidity sensor array arranged on the circuit breaker, and obtain the temperature information array and the humidity information array; A circuit characteristic prediction unit 12 is used to collect a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time, and perform circuit characteristic information prediction to obtain a predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes predicted circuit characteristic information at a future time node; The environmental impact analysis unit 13 is used to perform temperature impact analysis of future time nodes on the temperature information array according to the predicted circuit characteristic information sequence to obtain a predicted temperature information array, and perform humidity impact analysis of future time nodes based on the predicted temperature information array and the humidity information array to obtain a predicted humidity information array; A mechanical aging prediction unit 14 is used to collect recorded mechanical aging information of the circuit breaker, perform overheat aging prediction based on the predicted temperature information array, and obtain predicted mechanical aging information; A fault probability calculation unit 15 is used to perform a circuit breaker operation fault probability analysis based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information to obtain a circuit breaker fault probability; The failure risk assessment unit 16 is used to perform a circuit breaking protection demand analysis according to the predicted circuit characteristic information sequence to obtain a circuit breaking demand probability, and perform a failure monitoring analysis of the circuit breaker based on the prediction accuracy in combination with the circuit breaking failure probability to obtain a circuit breaking failure probability as a remote monitoring result.

[0058] Furthermore, the circuit characteristic prediction unit 12 includes the following execution steps: Collecting a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time; According to the historical operation data of the circuit in which the circuit breaker is located, a set of sample historical circuit feature information sequences before multiple historical time points is collected, and a set of sample predicted circuit feature information sequences after multiple historical time points is collected; Using the sample historical circuit feature information sequence set and the sample predicted circuit feature information sequence set to train a circuit feature information predictor; The circuit characteristic information predictor is used to perform circuit characteristic information prediction on the circuit characteristic information sequence to obtain predicted circuit characteristic information sequences at multiple time nodes within the future time node.

[0059] Furthermore, the environmental impact analysis unit 13 includes the following execution steps: Training a sensor information predictor based on historical sensor monitoring data of the circuit breaker, wherein the sensor information predictor includes a temperature information prediction branch and a humidity information prediction branch; The sensor information predictor is used to predict the temperature impact of future time nodes on the temperature information array and the prediction circuit characteristic information sequence to obtain a predicted temperature information array, and the humidity impact of future time nodes on the humidity information array and the prediction circuit characteristic information sequence is predicted to obtain a predicted humidity information array.

[0060] Furthermore, the environmental impact analysis unit 13 further includes the following execution steps: According to the historical sensor monitoring data of the circuit breaker, a sample circuit characteristic information sequence set is collected, and a sample basic temperature information array set and a sample basic humidity information array set at the starting time point are collected, and a sample predicted temperature information array set and a sample predicted humidity information array set at the predicted time node are collected; Using the sample circuit characteristic information sequence set, the sample basic temperature information array set, and the sample predicted temperature information array set to train a temperature information prediction branch; The sample circuit characteristic information sequence set, the sample basic humidity information array set and the sample predicted humidity information array set are used to predict the humidity information prediction branch; The temperature information prediction branch and the humidity information prediction branch are combined to obtain a sensor information predictor.

[0061] Furthermore, the mechanical aging prediction unit 14 includes the following execution steps: Collecting the mechanical aging degree of the circuit breaker during the last mechanical aging prediction as the recorded mechanical aging information; According to the mechanical aging test data of the circuit breaker, a sample temperature information array set and a sample mechanical aging degree set are collected to train a mechanical overheat aging analyzer; Based on the mechanical overheat aging analyzer, performing aging analysis on the recorded mechanical aging information to obtain an overheat aging degree; The predicted mechanical aging degree is calculated based on the recorded mechanical aging information and the overheat aging degree as the predicted mechanical aging information.

[0062] Further, the failure probability calculation unit 15 includes the following execution steps: According to the predicted temperature information array and the predicted humidity information array, average predicted temperature information and average predicted humidity information are calculated and obtained; According to the circuit breaker fault record data under different parameters, the sample average temperature information set, the sample average humidity information set, and the sample mechanical aging information set are collected, and according to the failure ratio of the circuit breaker under different parameters, the sample failure probability set is marked; Using the sample average temperature information set, the sample average humidity information set, the sample mechanical aging information set and the sample failure probability set to train a circuit breaker failure analyzer; Based on the circuit breaker fault analyzer, the circuit breaker failure probability is predicted based on the average predicted temperature information, the average predicted humidity information and the predicted mechanical aging information to obtain a circuit breaker failure probability.

[0063] Furthermore, the failure risk assessment unit 16 includes the following execution steps: According to the operation data records of the same type of circuits, a set of characteristic information sequences of the same type of circuits is collected, and according to the probability of the occurrence of circuit breaking protection requirements under different characteristic information sequences of the same type of circuits, a set of sample circuit breaking requirements probability is marked; Using the same family circuit feature information sequence set and the sample circuit breaking demand probability set, training a circuit breaking protection demand analyzer, performing circuit breaking protection demand analysis on the predicted circuit feature information sequence, and obtaining a circuit breaking demand probability; Testing the circuit breaker fault analyzer and the circuit breaker protection requirement analyzer to obtain the circuit breaker fault analysis accuracy and the circuit breaker requirement analysis accuracy; According to the circuit breaking demand probability and the circuit breaking failure probability, a basic circuit breaking failure probability is calculated; Based on the circuit breaking fault analysis accuracy and the circuit breaking demand analysis accuracy, a compensation calculation is performed on the basic circuit breaking failure probability to obtain a circuit breaking failure probability as a remote monitoring result.

[0064] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0065] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A remote monitoring method for circuit breakers based on multi-source sensor fusion, characterized in that: The method comprises: Based on the temperature sensor array and humidity sensor array arranged on the circuit breaker, the temperature information and humidity information of the circuit breaker are remotely monitored and collected to obtain the temperature information array and the humidity information array; Collecting a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time, and predicting the circuit characteristic information to obtain a predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes predicted circuit characteristic information at a future time node; According to the predicted circuit characteristic information sequence, the temperature information array is subjected to temperature influence analysis at future time nodes to obtain a predicted temperature information array, and based on the predicted temperature information array and the humidity information array, the humidity influence analysis at future time nodes is subjected to humidity influence analysis to obtain a predicted humidity information array; Collecting recorded mechanical aging information of the circuit breaker, performing overheat aging prediction based on the predicted temperature information array, and obtaining predicted mechanical aging information; Based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information, a circuit breaker operation failure probability analysis is performed to obtain a circuit breaker failure probability; A circuit breaking protection demand analysis is performed according to the predicted circuit characteristic information sequence to obtain a circuit breaking demand probability. In combination with the circuit breaking failure probability, based on the prediction accuracy, a failure monitoring analysis of the circuit breaker is performed to obtain a circuit breaking failure probability as a remote monitoring result.

2. The circuit breaker remote monitoring method based on multi-source sensor fusion according to claim 1 is characterized in that: Collecting a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time, and predicting the circuit characteristic information to obtain a predicted circuit characteristic information sequence, including: Collecting a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time; According to the historical operation data of the circuit in which the circuit breaker is located, a set of sample historical circuit feature information sequences before multiple historical time points is collected, and a set of sample predicted circuit feature information sequences after multiple historical time points is collected; Using the sample historical circuit feature information sequence set and the sample predicted circuit feature information sequence set to train a circuit feature information predictor; The circuit characteristic information predictor is used to perform circuit characteristic information prediction on the circuit characteristic information sequence to obtain predicted circuit characteristic information sequences at multiple time nodes within the future time node.

3. The circuit breaker remote monitoring method based on multi-source sensor fusion according to claim 1 is characterized in that: According to the predicted circuit characteristic information sequence, the temperature information array is subjected to temperature influence analysis at future time nodes to obtain a predicted temperature information array, and based on the predicted temperature information array and the humidity information array, the humidity influence analysis at future time nodes is performed to obtain a predicted humidity information array, including: Training a sensor information predictor based on historical sensor monitoring data of the circuit breaker, wherein the sensor information predictor includes a temperature information prediction branch and a humidity information prediction branch; The sensor information predictor is used to predict the temperature impact of future time nodes on the temperature information array and the prediction circuit characteristic information sequence to obtain a predicted temperature information array, and the humidity impact of future time nodes on the humidity information array and the prediction circuit characteristic information sequence is predicted to obtain a predicted humidity information array.

4. The circuit breaker remote monitoring method based on multi-source sensor fusion according to claim 3 is characterized in that: Based on the historical sensor monitoring data of the circuit breaker, the sensor information predictor is trained, including: According to the historical sensor monitoring data of the circuit breaker, a sample circuit characteristic information sequence set is collected, and a sample basic temperature information array set and a sample basic humidity information array set at the starting time point are collected, and a sample predicted temperature information array set and a sample predicted humidity information array set at the predicted time node are collected; Using the sample circuit characteristic information sequence set, the sample basic temperature information array set, and the sample predicted temperature information array set to train a temperature information prediction branch; The sample circuit characteristic information sequence set, the sample basic humidity information array set and the sample predicted humidity information array set are used to predict the humidity information prediction branch; The temperature information prediction branch and the humidity information prediction branch are combined to obtain a sensor information predictor.

5. The circuit breaker remote monitoring method based on multi-source sensor fusion according to claim 1 is characterized in that: Collecting the recorded mechanical aging information of the circuit breaker, performing overheat aging prediction based on the predicted temperature information array, and obtaining predicted mechanical aging information, including: Collecting the mechanical aging degree of the circuit breaker during the last mechanical aging prediction as the recorded mechanical aging information; According to the mechanical aging test data of the circuit breaker, a sample temperature information array set and a sample mechanical aging degree set are collected to train a mechanical overheat aging analyzer; Based on the mechanical overheat aging analyzer, performing aging analysis on the recorded mechanical aging information to obtain an overheat aging degree; The predicted mechanical aging degree is calculated based on the recorded mechanical aging information and the overheat aging degree as the predicted mechanical aging information.

6. The circuit breaker remote monitoring method based on multi-source sensor fusion according to claim 1, characterized in that: Based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information, a circuit breaker operation failure probability analysis is performed to obtain a circuit breaker failure probability, including: According to the predicted temperature information array and the predicted humidity information array, average predicted temperature information and average predicted humidity information are calculated and obtained; According to the circuit breaker fault record data under different parameters, the sample average temperature information set, the sample average humidity information set, and the sample mechanical aging information set are collected, and according to the failure ratio of the circuit breaker under different parameters, the sample failure probability set is marked; Using the sample average temperature information set, the sample average humidity information set, the sample mechanical aging information set and the sample failure probability set to train a circuit breaker failure analyzer; Based on the circuit breaker fault analyzer, the circuit breaker failure probability is predicted based on the average predicted temperature information, the average predicted humidity information and the predicted mechanical aging information to obtain a circuit breaker failure probability.

7. The circuit breaker remote monitoring method based on multi-source sensor fusion according to claim 6 is characterized in that: Performing a circuit breaker protection demand analysis according to the predicted circuit characteristic information sequence to obtain a circuit breaker demand probability, and performing a circuit breaker failure monitoring analysis based on the prediction accuracy in combination with the circuit breaker failure probability to obtain a circuit breaker failure probability, including: According to the operation data records of the same type of circuits, a set of characteristic information sequences of the same type of circuits is collected, and according to the probability of the occurrence of circuit breaking protection requirements under different characteristic information sequences of the same type of circuits, a set of sample circuit breaking requirements probability is marked; Using the same family circuit feature information sequence set and the sample circuit breaking demand probability set, training a circuit breaking protection demand analyzer, performing circuit breaking protection demand analysis on the predicted circuit feature information sequence, and obtaining a circuit breaking demand probability; Testing the circuit breaker fault analyzer and the circuit breaker protection requirement analyzer to obtain the circuit breaker fault analysis accuracy and the circuit breaker requirement analysis accuracy; According to the circuit breaking demand probability and the circuit breaking failure probability, a basic circuit breaking failure probability is calculated; Based on the circuit breaking fault analysis accuracy and the circuit breaking demand analysis accuracy, a compensation calculation is performed on the basic circuit breaking failure probability to obtain a circuit breaking failure probability as a remote monitoring result.

8. A circuit breaker remote monitoring device based on multi-source sensor fusion, characterized in that: The device for implementing the circuit breaker remote monitoring method based on multi-source sensor fusion according to any one of claims 1 to 7 comprises: A multi-source information acquisition unit, the multi-source information acquisition unit is used to remotely monitor and acquire temperature information and humidity information of the circuit breaker based on a temperature sensor array and a humidity sensor array arranged on the circuit breaker, and obtain a temperature information array and a humidity information array; A circuit characteristic prediction unit, the circuit characteristic prediction unit is used to collect a circuit characteristic information sequence of the circuit where the circuit breaker is located in the past time, and perform circuit characteristic information prediction to obtain a predicted circuit characteristic information sequence, wherein the predicted circuit characteristic information sequence includes predicted circuit characteristic information at a future time node; An environmental impact analysis unit, the environmental impact analysis unit is used to perform a temperature impact analysis of a future time node on the temperature information array according to the predicted circuit characteristic information sequence to obtain a predicted temperature information array, and perform a humidity impact analysis of a future time node based on the predicted temperature information array and the humidity information array to obtain a predicted humidity information array; A mechanical aging prediction unit, the mechanical aging prediction unit is used to collect recorded mechanical aging information of the circuit breaker, perform overheat aging prediction based on the predicted temperature information array, and obtain predicted mechanical aging information; A fault probability calculation unit, the fault probability calculation unit is used to perform circuit breaker operation fault probability analysis based on the predicted temperature information array, the predicted humidity information array and the predicted mechanical aging information to obtain a circuit breaker fault probability; A failure risk assessment unit, wherein the failure risk assessment unit is used to perform a circuit breaking protection demand analysis according to the predicted circuit characteristic information sequence to obtain a circuit breaking demand probability, and in combination with the circuit breaking failure probability, perform a failure monitoring analysis of the circuit breaker based on the prediction accuracy to obtain a circuit breaking failure probability as a remote monitoring result.

Images