A method for predicting the electrical life of circuit breaker contacts and providing intelligent alarms
By collecting data in real time at the circuit breaker contact points and utilizing an exponential degradation model and adaptive threshold method, combined with a multi-level early warning mechanism, accurate prediction and intelligent alarm of the electrical life of circuit breaker contacts are achieved. This solves the problem that existing technologies cannot meet the requirements for predicting the electrical life of circuit breaker contacts and improves the predictive maintenance capabilities of the equipment.
Patent Information
- Application Number
- CN202611346513.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-01
- Publication Date
- 2026-11-10
AI Technical Summary
Existing circuit breaker life prediction technologies cannot meet the specific requirements of circuit breaker contact electrical life and are insufficient to meet the predictive maintenance requirements of smart grids.
By using a sensor array installed at the circuit breaker contacts to collect contact temperature, contact resistance, and arcing parameters in real time, the electrical life is predicted using an exponential degradation model and an adaptive threshold method. Combined with a multi-level early warning mechanism and an intelligent alarm system, accurate prediction and intelligent alarm of contact electrical life can be achieved.
It enables accurate prediction of the electrical life of circuit breaker contacts, meets the power system's need for predictive maintenance of equipment, and improves operation and maintenance efficiency and equipment reliability.
Smart Images

Figure CN123066546A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system technology, and in particular to a method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm. Background Technology
[0002] With the rapid development of power systems, circuit breakers, as crucial protection and control devices, play a vital role in ensuring the reliability of the power grid through their safe and stable operation. Circuit breaker contacts, as key components for interrupting fault currents, are subjected to multiple aging factors such as arc erosion and mechanical wear during frequent operation. Their electrical life directly affects the circuit breaker's breaking performance and operational reliability. However, existing circuit breaker life prediction technologies either target the circuit breaker as a whole rather than specific contacts, or they are life prediction technologies for other fields. These technologies cannot meet the specific needs of predicting the electrical life of circuit breaker contacts and thus fail to meet the predictive maintenance requirements of smart grids. Summary of the Invention
[0003] In view of the above, the present invention aims to provide a method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm, so as to solve the aforementioned technical problems.
[0004] The technical solution adopted in this invention is as follows:
[0005] This invention provides a method for predicting the electrical life of circuit breaker contacts and for intelligent alarm, including:
[0006] Raw data is obtained by real-time acquisition of contact temperature, contact resistance, and arcing parameters by a sensor array installed at the circuit breaker contacts.
[0007] The collected raw data were preprocessed, and life degradation characteristic parameters related to contact electrical life degradation were extracted.
[0008] Based on the aforementioned life degradation characteristic parameters, the current cumulative value of electrical life consumption of the contact is calculated using an exponential degradation model. At the same time, an adaptive threshold method is used to perform trend analysis on the electrical life consumption sequence, predict the electrical life consumption trend of the contact within a set number of future operations, and determine the percentage of remaining electrical life of the contact based on the prediction results.
[0009] The warning threshold is dynamically adjusted based on the electrical lifetime prediction results. When the cumulative electrical lifetime consumption exceeds the preset threshold, the corresponding level of warning signal is triggered.
[0010] The alarm device is activated according to the warning level, and the warning event is automatically recorded to local storage and uploaded to the operation and maintenance platform.
[0011] Optionally, the sensor group includes a temperature sensor, a contact resistance sensor, and an arc voltage sampling circuit;
[0012] The raw data includes contact temperature, contact resistance, and circuit breaker arcing parameters;
[0013] The temperature sensor is used to measure the working temperature of the contacts, the contact resistance sensor is used to measure the contact resistance value when the contacts are closed, and the arc voltage sampling circuit is used to collect the arc voltage signal during the opening process and extract the arc time and arc energy.
[0014] Optionally, the contact resistance sensor uses a four-wire measurement method to measure the contact resistance value when the contacts are closed.
[0015] Optionally, the arc voltage sampling circuit extracts the arc energy parameters by integrating the arc voltage and the current.
[0016] Optionally, the contact resistance sensor uses a dynamic threshold method to identify the initial contact resistance value at the moment the contact closes.
[0017] Optionally, the lifetime degradation characteristic parameters include resistance characteristics, temperature characteristics, and arcing energy characteristics;
[0018] The contact resistance characteristics include the initial contact resistance value and the resistance growth rate; the temperature characteristics include the contact operating temperature and the temperature gradient; and the arcing energy characteristics include the arcing energy value for each opening operation and the cumulative arcing energy.
[0019] Optionally, the exponential degradation model uses an exponentially weighted moving average method to calculate the cumulative value of contact electrical life consumption, and the model parameters are adaptively adjusted based on the historical operating data of the contacts.
[0020] Optionally, the threshold setting of the adaptive threshold method is dynamically adjusted according to the statistical characteristics of the electrical lifetime consumption sequence, and the electrical lifetime consumption trend within a set number of future cycles is predicted by performing linear regression on the electrical lifetime consumption sequence.
[0021] Optionally, the warning threshold can be dynamically adjusted based on the electrical lifetime prediction results. When the cumulative electrical lifetime consumption exceeds a preset threshold, a warning signal of the corresponding level is triggered, including:
[0022] The warning thresholds are divided into multiple levels based on the remaining battery life. When the remaining battery life is greater than the first preset threshold, it is determined to be in a normal state. When the remaining battery life is in the range between the second preset threshold and the first preset threshold, a notice warning is triggered. When the remaining battery life is in the range between the third preset threshold and the second preset threshold, a warning is triggered. When the remaining battery life is less than the third preset threshold, an emergency warning is triggered.
[0023] Optionally, the circuit breaker contact electrical life prediction and intelligent alarm method also includes an alarm delay debounce mechanism. The alarm delay setting range is adjusted according to actual needs, and an alarm event is determined only when the continuous triggering condition lasts for the entire delay period.
[0024] The above-described solution of the present invention has at least the following beneficial effects:
[0025] The above-described solution of the present invention collects contact temperature, contact resistance, and arcing parameters in real time using a sensor array installed at the circuit breaker contacts to obtain raw data. The collected raw data is preprocessed, and life degradation characteristic parameters related to contact electrical life degradation are extracted. Based on these life degradation characteristic parameters, an exponential degradation model is used to calculate the current cumulative value of contact electrical life consumption. Simultaneously, an adaptive threshold method is used to perform trend analysis on the electrical life consumption sequence, predicting the electrical life consumption trend of the contacts within a set number of future operations, and determining the percentage of remaining contact electrical life based on the prediction results. The warning threshold is dynamically adjusted based on the electrical life prediction results. When the cumulative value of electrical life consumption exceeds a preset threshold, a warning signal of the corresponding level is triggered. The corresponding alarm device is activated according to the warning level, and the warning event is automatically recorded in local storage and simultaneously uploaded to the operation and maintenance platform. The solution of the present invention can achieve intelligent early warning of contact electrical life, meeting the needs of power systems for predictive maintenance of equipment. Attached Figure Description
[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings, wherein:
[0027] Figure 1 A flowchart of a circuit breaker contact electrical life prediction and intelligent alarm method provided in an embodiment of the present invention.
[0028] Figure 2 This is a schematic diagram of the architecture of the circuit breaker contact electrical life prediction and intelligent alarm system provided in an embodiment of the present invention. Detailed Implementation
[0029] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.
[0030] This invention proposes an embodiment of a method for predicting the electrical life of circuit breaker contacts and providing intelligent alarms. Specifically, as follows: Figure 1 As shown, it includes:
[0031] Step 11: Collect contact temperature, contact resistance and arcing parameters in real time by using a sensor group installed at the circuit breaker contact to obtain raw data;
[0032] Step 12: Preprocess the collected raw data and extract the lifetime degradation characteristic parameters related to the electrical lifetime degradation of the contacts;
[0033] Step 13: Based on the life degradation characteristic parameters, the current cumulative value of electrical life consumption of the contact is calculated using an exponential degradation model. At the same time, the trend analysis of the electrical life consumption sequence is performed using an adaptive threshold method to predict the electrical life consumption trend of the contact within a set number of future operations. The remaining electrical life percentage of the contact is determined based on the prediction results.
[0034] Step 14: Dynamically adjust the warning threshold based on the electrical lifetime prediction results. When the cumulative electrical lifetime consumption exceeds the preset threshold, trigger the corresponding level of warning signal.
[0035] Step 15: Drive the corresponding alarm device according to the warning level, and automatically record the warning event to local storage and upload it to the operation and maintenance platform.
[0036] In this embodiment, the circuit breaker contact electrical life prediction and intelligent alarm method is applied to the circuit breaker contact electrical life prediction and intelligent alarm system, such as... Figure 2 As shown, the system architecture in this embodiment comprises four layers: a data acquisition layer, a data processing layer, a predictive analysis layer, and an alarm output layer. The data acquisition layer is responsible for acquiring, conditioning, and digitizing sensor signals, serving as the system's front-end sensing component. The data processing layer is responsible for preprocessing raw data and calculating feature parameters, implementing data filtering, noise reduction, and feature extraction functions. The predictive analysis layer is responsible for calculating the electrical life prediction model and determining the warning threshold, achieving accurate prediction of contact electrical life. The alarm output layer is responsible for generating and driving alarm signals, enabling multi-level alarm linkage. These four layers collaborate to complete the full functionality of contact electrical life prediction and intelligent alarm.
[0037] The acquisition of contact parameters in step 11 provides the raw data foundation for the entire system, and its acquisition accuracy and real-time performance directly affect the accuracy of subsequent electrical life prediction.
[0038] In this embodiment, the sensor group includes three core sensing units: a temperature sensor, a contact resistance sensor, and an arc voltage sampling circuit. The temperature sensor, employing a high-precision thermocouple or infrared temperature sensing element, is installed near the contact area of the moving and stationary contacts of the circuit breaker to monitor temperature changes during contact operation in real time. The temperature sensor's measurement accuracy is set to ±0.5℃ to ±1℃, with a response time not exceeding 10ms, ensuring the ability to capture instantaneous temperature fluctuations during contact opening and closing. Since ambient temperature interferes with contact temperature measurement, the temperature sensor module is equipped with a temperature compensation algorithm. This algorithm, based on ambient temperature data collected by the ambient temperature sensor, corrects the contact temperature measurement value in real time, eliminating the influence of ambient temperature drift on the measurement results. The compensation algorithm uses a linear correction model or a polynomial correction model, with specific parameters determined based on measured data from the circuit breaker installation site.
[0039] A contact resistance sensor is used to measure the contact resistance value of circuit breaker contacts in the closed state. Contact resistance is a key parameter reflecting the contact state and degradation degree of the contact surface. In this embodiment, the contact resistance sensor adopts a four-wire measurement method. This method effectively eliminates the influence of lead resistance on measurement accuracy through two independent current loops and voltage measurement loops. Specifically, one pair of wires is used to apply a constant measurement current to the contacts, and the other pair of wires is used to collect the voltage drop across the contacts. Since the input impedance of the voltage measurement loop is extremely high, almost no current flows through the voltage measurement wires, so the lead resistance does not generate a voltage drop, thereby achieving high-precision contact resistance measurement. The measurement current is set to a preset current value, usually selected in the range of 10mA to 100mA. This current value can meet the signal-to-noise ratio requirements of contact resistance measurement and avoid the interference of the measurement results caused by the thermal effect of excessive measurement current causing the contacts to heat up. In the actual measurement process, the measurement current is applied after the contacts are closed and stabilized, and the duration is controlled within several hundred milliseconds. Then the measurement current is cut off to ensure that the measurement process does not affect the normal working state of the contacts.
[0040] The arc voltage sampling circuit is a key module for acquiring the arc voltage signal during the circuit breaker's opening operation. Its sampling accuracy and timing directly determine the quality of the extracted arc energy parameters. The sampling rate of the arc voltage sampling circuit reaches a preset frequency, which is determined based on the circuit breaker's breaking characteristics and the arc duration, typically set to the range of 100kHz to 1MHz to ensure complete capture of the transient changes in arc voltage. The arc voltage sampling circuit starts sampling immediately upon detecting the opening operation trigger signal, completing arc voltage signal acquisition within a preset time period after opening. This preset time period covers the entire arcing process, and its duration is determined based on the circuit breaker's rated voltage and breaking current rating, typically set to the range of 10ms to 100ms. The arc voltage signal acquired by the arc voltage sampling circuit is low-pass filtered and signal conditioned before being sent to an analog-to-digital converter for digital processing. The extraction of arc energy is achieved through the integral calculation of arc voltage and current, using the following formula:
[0041] ;
[0042] in, The arc energy of a single circuit breaker trip is expressed in joules (J). It represents the change of arc voltage over time, and its unit is volt (V). It is a function representing the change of arc current over time, and its unit is ampere (A). and These represent the arc start and end times, respectively. The arc voltage sampling circuit synchronously acquires the arc voltage and arc current signals. After undergoing the same time-base processing, both signals are synchronously integrated in the digital processing unit to obtain the arc energy value for a single circuit breaker trip. The sampling circuit also records the arc duration, a parameter that serves as an important component of the arc energy characteristics and participates in subsequent electrical lifetime prediction calculations.
[0043] Furthermore, step 12 preprocesses and performs feature calculations on the raw data collected in step 11 to extract core feature parameters related to the degradation of contact electrical life. Raw data preprocessing includes three stages: data filtering, denoising, and outlier handling. Data filtering employs Kalman filtering or moving average filtering to eliminate the influence of sensor measurement noise and power frequency interference on the data. Denoising targets high-frequency noise components in the arc voltage sampling signal, using wavelet transform or digital filtering for frequency domain processing. Outlier handling assesses the reasonableness of the sensor-collected data, eliminating abnormal data points caused by sensor malfunction or electromagnetic interference. Outlier criteria include data exceeding the sensor's measurement range, data change rate exceeding the physically achievable range, and severe mismatch with other relevant parameters.
[0044] Contact resistance feature extraction comprises two sub-steps: initial contact resistance value extraction and resistance growth rate calculation. The initial contact resistance value extraction employs a dynamic threshold method. This method sets a dynamic threshold on the contact resistance change curve at the instant the contact closes, detects abrupt resistance changes during the process from pre-contact to full contact, and identifies the resistance value corresponding to these abrupt changes as the initial contact resistance value. Specifically, the system records a continuous data sequence of contact resistance changes with displacement during contact closure, uses differential calculations to detect the peak position of the resistance change rate, and the resistance value corresponding to this peak position is the initial contact resistance value. The dynamic threshold is set based on the contact characteristics of the contact material, typically selecting 50% to 80% of the peak change rate as the threshold determination point. The resistance growth rate calculation combines data from multiple consecutive operations, taking the contact resistance measurements corresponding to the most recent N opening and closing operations, and calculating the resistance growth trend with the number of operations using linear regression or exponential fitting methods to obtain the resistance growth rate parameter. This growth rate parameter reflects the degree of arc erosion and material migration on the contact surface and is an important basis for predicting the remaining electrical life of the contact.
[0045] Temperature feature extraction includes two parameters: contact operating temperature and temperature gradient. The contact operating temperature is obtained directly from the measurement value of the temperature sensor, after correction by a temperature compensation algorithm. The temperature gradient is defined as the rate of change of the contact temperature relative to time, and is obtained by performing a difference operation on the temperature measurement sequence. The calculation formula is as follows:
[0046] ;
[0047] in, This represents the temperature gradient, with units of °C / s. and These represent the contact temperature values at two consecutive sampling times; This indicates the sampling time interval. The temperature gradient parameter reflects the heat accumulation and heat dissipation characteristics of the contact during operation. When the contact resistance increases or the arcing energy increases, the contact temperature gradient will show an upward trend.
[0048] Arc energy feature extraction uses the arc energy value of each opening operation as the basic feature, and calculates the cumulative arc energy sequence as needed. The single arc energy value is directly obtained by the integration operation of the arc voltage sampling circuit. The cumulative arc energy sequence is calculated by accumulating and summing the arc energy of each historical opening operation. This sequence reflects the total arc erosion experienced by the contact during long-term operation. The calculation of cumulative arc energy adopts a recursive form:
[0049] ;
[0050] in, This represents the cumulative arc energy after the k-th tripping operation; This represents the cumulative arc energy after the (k-1)th tripping operation; This represents the arcing energy value during the k-th tripping operation. The cumulative arcing energy sequence serves as an important input parameter for the electrical lifetime prediction model, and together with the contact resistance growth rate and temperature characteristics, it constitutes a multidimensional degradation feature space.
[0051] In step 13, the electrical lifetime prediction is based on the lifetime degradation characteristic parameters extracted in step 12. An exponential degradation model is used to calculate the current cumulative electrical lifetime consumption of the contact, and an adaptive threshold method and linear regression prediction method are combined to analyze the future electrical lifetime consumption trend. This step is the core technical component of the entire method, and its prediction accuracy directly determines the reliability of the early warning decision.
[0052] The exponential degradation model is constructed based on the physical mechanism of contact electrical life degradation. During each opening operation, the contact is subjected to arc erosion and mechanical wear, and its degradation degree has a non-linear relationship with the number of operations and the breaking conditions. This embodiment uses an exponential degradation model to describe the cumulative process of contact electrical life consumption. The model uses an exponentially weighted moving average method to calculate the cumulative value of contact electrical life consumption. The exponentially weighted moving average method performs weighted averaging of electrical life consumption from each historical operation, with recent data having a larger weight and older data having a weight that decreases exponentially. The formula for calculating the cumulative value of electrical life consumption is:
[0053] ;
[0054] in, This represents the cumulative electrical lifetime consumption after the k-th operation; This represents the electrical lifetime consumption corresponding to the k-th operation; This represents the exponential weighting coefficient, a constant ranging from 0 to 1.
[0055] Electrical life consumption The overall degradation level is calculated by comprehensively considering multiple degradation indices, including contact resistance growth rate, contact temperature, and arc energy. The specific calculation method involves first normalizing each degradation index, and then using a weighted summation or Euclidean distance method to calculate the comprehensive degradation level. (Exponential weighting coefficient) The value is adaptively adjusted based on the historical operating data of the contact. The initial value is set in the range of 0.1 to 0.3. When the contact degradation enters the accelerated phase, The value is automatically increased to give more weight to recent data in the prediction. The adaptive adjustment of model parameters is achieved through least squares or recursive least squares methods. The system continuously tracks the error between the model's predicted values and the actual observed values, and triggers parameter updates when the error exceeds a preset threshold.
[0056] The adaptive threshold method dynamically adjusts the threshold setting based on the statistical characteristics of the electrical lifetime consumption sequence. Statistical analysis is performed on the electrical lifetime consumption sequence to calculate its mean, standard deviation, and rate of change, and the warning threshold is automatically adjusted based on these statistical characteristics. This method effectively adapts to individual differences in contacts of different circuit breaker models and changes in the operating environment, significantly improving the generalization ability of the prediction model. In specific implementation, a sliding time window is maintained, with the window length set to the electrical lifetime consumption data corresponding to the most recent 50 to 100 operations. The mean and standard deviation of the data within the window serve as the baseline level for the current electrical lifetime consumption. The warning threshold is dynamically adjusted based on this baseline level, and the threshold calculation formula is as follows:
[0057] ;
[0058] in, This represents the dynamic warning threshold at the k-th operation. This represents the mean of the electrical lifetime consumption sequence within the sliding window; The standard deviation of the electrical lifetime consumption sequence within the sliding window; This represents the threshold coefficient, which is set according to the warning level and reliability requirements. The value range is usually from 2 to 4, when When the value is larger, the warning sensitivity decreases. Smaller values improve warning sensitivity. The statistical data within the sliding window is continuously updated based on actual operational data to achieve adaptive threshold tracking.
[0059] Trend prediction uses linear regression analysis on the electrical life consumption sequence to predict the electrical life consumption trend of the contacts within a set number of future operations. The linear regression model takes the following form:
[0060] in, This represents the cumulative electrical lifetime consumption predicted from the current k-th operation to the (k+n)-th operation; n represents the prediction step size, i.e., the number of future operations; a represents the linear regression slope, reflecting the growth rate of electrical lifetime consumption. The regression slope a is obtained by least-squares fitting of the electrical lifetime consumption sequence. The length of the fitting data window is determined according to the prediction accuracy requirements, usually selecting data from the most recent 20 to 50 operations. The prediction step size n is set according to maintenance requirements, usually set to the range of 10 to 50 operations. The percentage of remaining electrical lifetime of the contact is calculated based on the prediction results, using the following formula:
[0061] ;
[0062] in, Indicates the percentage of remaining electrical life of the contacts; The rated electrical life of the contacts is usually expressed as the rated number of operating cycles or the rated cumulative arc energy. This represents the cumulative electrical lifetime consumption predicted after n future operations. The remaining electrical lifetime percentage is a core indicator for determining contact condition and triggering early warnings.
[0063] In step 14, the warning threshold is dynamically adjusted based on the electrical lifetime prediction results from step 13. When the accumulated electrical lifetime consumption exceeds the preset threshold, a warning signal of the corresponding level is triggered. This embodiment establishes a three-level warning system, dividing the warning levels into four states: normal, attention, warning, and emergency, corresponding to different electrical lifetime remaining threshold ranges.
[0064] The first warning threshold is the attention warning threshold. When the remaining electrical life is greater than this threshold, it is considered a normal state, indicating that the contact is in good operating condition and no maintenance measures are required. The attention warning threshold is set considering the early characteristics of contact degradation. This level of warning is triggered when the cumulative electrical life consumption begins to show an upward trend but has not yet reached a significant degradation level. After the attention warning is triggered, the system outputs a prompt message to remind maintenance personnel to pay attention to the trend of contact status changes, but no immediate action is required.
[0065] The second warning threshold is the alert threshold. A warning is triggered when the remaining electrical life falls within the range of the second warning threshold to the first warning threshold, indicating that the contact has entered a significant degradation stage and a maintenance plan needs to be prepared. The width of the interval between the alert and the warning threshold is determined based on the contact degradation characteristics and the operation and maintenance strategy, and is usually set as a certain percentage of the total electrical life. After the warning is triggered, the system automatically generates a maintenance pre-warning work order, prompting maintenance personnel to begin preparing for contact repair or replacement work.
[0066] The third warning threshold is the emergency warning threshold. A warning is triggered when the remaining electrical life is between the third and second warning thresholds, indicating severe contact degradation and requiring immediate maintenance. Upon triggering an emergency warning, the system immediately notifies relevant maintenance personnel, shortening maintenance response time. An emergency warning is triggered when the remaining electrical life is less than the third warning threshold, indicating that the contact is nearing the end of its lifespan and immediate power outage maintenance or contact replacement is necessary. After an emergency warning is triggered, the system continuously outputs alarm signals until maintenance personnel confirm and take appropriate measures.
[0067] Furthermore, it includes an alarm delay debouncing mechanism. This mechanism improves the reliability of early warning judgments and avoids false alarms caused by instantaneous data fluctuations or measurement noise. The alarm delay setting range can be adjusted according to actual needs, typically ranging from 1 second to 60 seconds. The system only determines a true alarm event and outputs an alarm signal when the continuous triggering condition lasts for the entire delay period. If the triggering condition disappears within the delay period, the alarm judgment is canceled. The delay debouncing mechanism effectively filters out the impact of occasional data spikes and instantaneous interference on early warning judgments, improving the stability and reliability of the early warning system.
[0068] In step 15, the intelligent alarm output drives the corresponding alarm device according to the warning level, realizing multi-level alarm linkage function. The alarm devices in this embodiment include three types: audible and visual alarms, on-site indicator lights, and a background monitoring system. Different warning levels correspond to different alarm driving strategies to ensure the effective transmission of alarm information.
[0069] The audible and visual alarm system employs a combined design, comprising an audible alarm unit and a visual alarm unit. The audible alarm unit emits alarm sounds of different frequencies and rhythms to distinguish warning levels: intermittent, short beeps indicate a warning, a continuous, medium-frequency warning, and a high-frequency, rapid alarm for an emergency. The visual alarm unit uses multi-color LED indicators or an LCD display: green indicates normal operation, yellow indicates a warning, orange indicates a hazard, and red or flashing indicates an emergency. The audible and visual alarm system's drive circuit receives alarm signals from the system control unit and controls the alarm's operation based on the warning level parameters.
[0070] Field indicator lights are installed on the circuit breaker body or switchgear panel, providing intuitive display of contact status. The indicator lights use a combination of three or four colors of LEDs, corresponding to different warning levels. The indicator lights operate in various modes, including constant on, flashing, and alternating flashing, conveying contact status information through different light combinations. The control signals for the field indicator lights are sent by the system control unit via hardwiring or a communication interface, ensuring the real-time nature and reliability of alarm information.
[0071] The background monitoring system receives early warning information and contact status data output by the system, providing a visual monitoring interface and historical data query function. The background monitoring system consists of three components: a data server, a monitoring workstation, and a communication network. The data server stores contact electrical life-related data, including raw acquired data, characteristic parameters, prediction results, and alarm records. The monitoring workstation runs monitoring software, displaying contact status parameters and early warning information in real time, and providing an interface for alarm confirmation, parameter setting, and historical data query. The communication network uses industrial Ethernet or a dedicated communication protocol to achieve data interaction between the field acquisition units and the background monitoring system.
[0072] Warning events are automatically recorded to local storage using non-volatile memory. The recorded information includes the warning occurrence time, warning level, trigger parameter values, contact status snapshot, and subsequent processing records. Warning events are also simultaneously uploaded to the operations and maintenance (O&M) platform. The O&M platform manages and analyzes the received warning information, automatically generating O&M work orders and dispatching them to relevant maintenance personnel. The O&M work orders include contact location information, warning level, processing suggestions, and time limits, guiding O&M personnel to carry out targeted maintenance work.
[0073] It also includes a historical data deep learning mechanism. When the system accumulates a preset threshold of historical data records, it automatically activates a recurrent neural network model to optimize and update the prediction model parameters to improve prediction accuracy. This mechanism enables the prediction model to continuously learn and self-optimize, adapting to the performance changes and degradation patterns of the contacts during long-term operation.
[0074] The input to the recurrent neural network model is historical electrical lifetime consumption sequence data, and the output is the optimized model parameters. The model employs a long short-term memory network or gated recurrent unit structure, which can effectively capture long-term dependencies and temporal features in the electrical lifetime consumption sequence. The model training data includes feature parameters and corresponding electrical lifetime consumption for each historical tripping operation; the data window length is determined based on model training performance and computational resources. Model training uses a backpropagation time-based algorithm, and hyperparameters such as the learning rate, number of hidden layer nodes, and number of recurrent layers are optimized and adjusted based on the validation set performance.
[0075] The workflow of a deep learning mechanism includes three stages: data accumulation, model training, and parameter update. In the data accumulation stage, the system continuously collects and stores operational data. When the number of historical data records reaches a preset threshold, model training is triggered. The preset threshold is set considering the minimum amount of data required for model training and system storage resources, typically between 1000 and 5000 records. The model training stage employs offline training, utilizing the accumulated historical data to train and optimize the recurrent neural network model, obtaining the optimal model parameters after training. In the parameter update stage, the optimal parameters obtained from training are deployed to the online prediction model, replacing the original model parameters and achieving iterative optimization of the prediction model. The deep learning mechanism executes periodically or on demand, continuously improving the accuracy and adaptability of the prediction model.
[0076] In practical applications, the circuit breaker contact electrical life prediction and intelligent alarm system of this invention can be deployed in power system substations or distribution network switching stations with voltage levels ranging from 35kV to 500kV. The system's workflow is as follows: Each time the circuit breaker completes a tripping operation, the sensor group collects raw data such as contact temperature, contact resistance, and arc voltage signals. The data acquisition time window covers a period of 50ms to 100ms before and after the tripping operation. The collected raw data is amplified, filtered, and converted from analog to digital by a signal conditioning circuit before being sent to the data processing unit for feature extraction. The data processing unit calculates the initial contact resistance value using a dynamic threshold method, calculates the temperature gradient using differential calculation, and calculates the arcing energy value using integral calculation. After normalization, the feature parameters are input into the electrical life prediction model. The prediction model uses an exponentially weighted moving average method to calculate the cumulative electrical life consumption value, an adaptive threshold method to calculate the dynamic early warning threshold, and a linear regression method to predict the future electrical life consumption trend. The prediction results are judged based on a three-level early warning threshold to determine the current early warning level. If the warning level is "Attention" or higher, the system will activate the audible and visual alarms and on-site indicator lights to output alarm signals. Simultaneously, the warning information will be stored locally and uploaded to the backend monitoring system and the maintenance platform. Maintenance personnel will then plan contact repairs based on the warning information, completing maintenance work before the contacts reach an emergency warning state to prevent power outages due to contact failure.
[0077] Regarding specific parameter configurations, a typical application example of this embodiment is as follows: The temperature sensor uses a type K thermocouple with a measurement accuracy of ±1℃, a response time of 5ms, and a measurement range of 0℃ to 300℃. The contact resistance sensor uses a four-wire measurement circuit with a measurement current of 50mA, a measurement accuracy of ±0.5μΩ, and a resolution of 0.1μΩ. The arc voltage sampling circuit has a sampling rate of 500kHz, a sampling time of 20ms after circuit breaker tripping, and an arc energy calculation error of less than 5%. The initial value of the exponential weighting coefficient of the exponential degradation model is 0.2, and the parameter update cycle is once a week. The sliding window length of the adaptive threshold method is 50 operations, and the threshold coefficient is 3. The fitting window length of the linear regression prediction is 30 operations, and the prediction step size is 20 operations. The three-level warning thresholds are set as follows: the attention warning threshold corresponds to 85% of the remaining electrical life, the warning threshold corresponds to 70% of the remaining electrical life, and the emergency warning threshold corresponds to 50% of the remaining electrical life. The alarm delay is set to 10 seconds. The trigger threshold for the deep learning mechanism is 2000 historical records, and the model training cycle is once a month.
[0078] Regarding data storage format, this embodiment employs a structured data storage format to store various parameters and prediction results. The raw contact data storage format includes fields such as timestamp, circuit breaker number, contact number, temperature value, contact resistance value, arc voltage sampling sequence, and arc energy value. The feature parameter storage format includes fields such as timestamp, feature number, feature name, and feature value. The prediction result storage format includes fields such as timestamp, prediction model version, cumulative electrical lifetime consumption, remaining electrical lifetime percentage, prediction error, and confidence interval. The alarm record storage format includes fields such as timestamp, alarm level, trigger threshold, contact status snapshot, processing status, and processing time. All types of data are stored using a time-series database or relational database, supporting historical data querying, statistical analysis, and big data analysis applications.
[0079] Regarding communication interfaces, this embodiment includes analog input interfaces, digital input / output interfaces, serial communication interfaces, and Ethernet communication interfaces. The analog input interfaces receive analog voltage or current signals from temperature sensors and are equipped with signal conditioning circuits and analog-to-digital converters. The digital input / output interfaces receive circuit breaker opening / closing status signals and drive alarm devices; these interfaces support the acquisition and output of switch signals. The serial communication interfaces use RS485 or RS232 protocols for data interaction with field intelligent devices. The Ethernet communication interfaces support the TCP / IP protocol for data communication with the backend monitoring system. The configuration parameters of the communication interfaces include baud rate, data bits, stop bits, parity check, and communication protocol, which are configured according to the requirements of the field devices and the system.
[0080] The circuit breaker contact electrical life prediction and intelligent alarm method in this embodiment comprehensively monitors key degradation indicators such as contact temperature, contact resistance and arcing energy through multi-dimensional feature parameters. Combined with the exponential degradation model and adaptive threshold method, it can achieve accurate prediction of contact electrical life, and the prediction error is controlled within a preset range.
[0081] The warning threshold is dynamically adjusted based on the statistical characteristics of the electrical life consumption sequence, which effectively adapts to the individual differences of contacts of different types of circuit breakers and changes in the operating environment, and significantly improves the generalization ability of the prediction model.
[0082] By performing linear regression analysis on the electrical life consumption sequence, the electrical life consumption trend within a set number of operations can be predicted in advance, allowing maintenance personnel sufficient preparation time for maintenance. The advance warning amount meets the preset number of operations requirements. A three-level early warning mechanism of normal, attention, warning, and emergency is established to achieve full coverage from status reminders to emergency alarms, ensuring that contact degradation states of different severity can be responded to in a timely manner.
[0083] It supports multi-level linkage between audible and visual alarms, on-site indicator lights, and background monitoring systems, and automatically generates maintenance work orders containing contact electrical life status and early warning information to improve maintenance efficiency. After historical data accumulation reaches a preset threshold, it automatically activates a recurrent neural network model to optimize prediction parameters, enabling the prediction model to continuously learn the contact degradation pattern and adapt to changes in equipment performance during long-term operation.
[0084] An embodiment of the present invention also provides a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described in the above embodiments. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.
[0085] In this embodiment of the invention, a computer-readable storage medium is also provided, storing instructions that, when executed on a computer, cause the computer to perform the method described in the above embodiments. All implementations of the methods described in the above embodiments are applicable to this embodiment and can achieve the same technical effect.
[0086] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.
[0087] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm, characterized in that, include: Raw data is obtained by real-time acquisition of contact temperature, contact resistance, and arcing parameters by a sensor array installed at the circuit breaker contacts. The collected raw data were preprocessed, and life degradation characteristic parameters related to contact electrical life degradation were extracted. Based on the aforementioned life degradation characteristic parameters, the current cumulative value of electrical life consumption of the contact is calculated using an exponential degradation model. At the same time, an adaptive threshold method is used to perform trend analysis on the electrical life consumption sequence, predict the electrical life consumption trend of the contact within a set number of future operations, and determine the percentage of remaining electrical life of the contact based on the prediction results. The warning threshold is dynamically adjusted based on the electrical lifetime prediction results. When the cumulative electrical lifetime consumption exceeds the preset threshold, the corresponding level of warning signal is triggered. The alarm device is activated according to the warning level, and the warning event is automatically recorded to local storage and uploaded to the operation and maintenance platform.
2. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 1, characterized in that, The sensor group includes a temperature sensor, a contact resistance sensor, and an arc voltage sampling circuit; The raw data includes contact temperature, contact resistance, and circuit breaker arcing parameters; The temperature sensor is used to measure the working temperature of the contacts, the contact resistance sensor is used to measure the contact resistance value when the contacts are closed, and the arc voltage sampling circuit is used to collect the arc voltage signal during the opening process and extract the arc time and arc energy.
3. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 2, characterized in that, The contact resistance sensor uses a four-wire measurement method to measure the contact resistance value when the contacts are closed.
4. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 2, characterized in that, The arc voltage sampling circuit extracts the arc energy parameters by integrating the arc voltage and the current.
5. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 2, characterized in that, The contact resistance sensor uses a dynamic threshold method to identify the initial contact resistance value at the moment the contact closes.
6. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 1, characterized in that, The lifetime degradation characteristic parameters include resistance characteristics, temperature characteristics, and arc energy characteristics; The contact resistance characteristics include the initial contact resistance value and the resistance growth rate; the temperature characteristics include the contact operating temperature and the temperature gradient; and the arcing energy characteristics include the arcing energy value for each opening operation and the cumulative arcing energy.
7. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 1, characterized in that, The exponential degradation model uses an exponentially weighted moving average method to calculate the cumulative value of contact electrical life consumption, and the model parameters are adaptively adjusted based on the historical operating data of the contacts.
8. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 1, characterized in that, The threshold setting of the adaptive threshold method is dynamically adjusted according to the statistical characteristics of the electrical lifetime consumption sequence, and the electrical lifetime consumption trend within a set number of future cycles is predicted by performing linear regression on the electrical lifetime consumption sequence.
9. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 1, characterized in that, The warning threshold is dynamically adjusted based on the electrical lifetime prediction results. When the cumulative electrical lifetime consumption exceeds the preset threshold, a warning signal of the corresponding level is triggered, including: The warning thresholds are divided into multiple levels based on the remaining battery life. When the remaining battery life is greater than the first preset threshold, it is determined to be in a normal state. When the remaining battery life is in the range between the second preset threshold and the first preset threshold, a notice warning is triggered. When the remaining battery life is in the range between the third preset threshold and the second preset threshold, a warning is triggered. When the remaining battery life is less than the third preset threshold, an emergency warning is triggered.
10. The method for predicting the electrical life of circuit breaker contacts and providing intelligent alarm according to claim 1, characterized in that, It also includes an alarm delay debouncing mechanism. The alarm delay setting range is adjusted according to actual needs. An alarm event is determined only when the continuous triggering condition lasts for the entire delay period.
