A circuit breaker current acquisition processing method and system

Abstract

The application discloses a kind of circuit breaker current acquisition processing method and system, belong to electric power equipment protection technical field, comprising: acquisition original analog current signal and multidimensional state parameter, preprocessed and packaged as digital current data packet with state label;Rely on the two-dimensional correction coefficient table of temperature zone, current amplitude layered calibration, complete three-phase current and zero sequence current amplitude, phase temperature drift compensation;Through three-phase synthesis zero sequence and actual measurement zero sequence amplitude, phase double threshold mutual check, output effective current data;Firstly meet low load condition when extracting three-phase direct current component mean as baseline bias, construct multi-source information fusion model, and adaptive optimization protection action criterion;While accessing disconnecting switch sub-position interlocking signal, forced current sampling is zero when breaking and additional interlocking mark, closing output check current sequence;The application suppresses temperature drift measurement error, realizes zero sequence double-way check to prevent sampling distortion, and reduces false rejection by dynamic optimization protection action criterion of multiple state parameters.

Description

Technical Field

[0001] This invention belongs to the field of power equipment protection technology, specifically a circuit breaker current acquisition and processing method and system. Background Technology

[0002] High-voltage circuit breakers rely on current acquisition data for fault protection, and the accuracy of this acquisition directly determines the reliability of the protection operation. Existing acquisition schemes have several shortcomings: First, current transformer sampling is susceptible to ambient temperature, resulting in amplitude and phase drift errors. Furthermore, the schemes fail to differentiate between the deviation characteristics of different current amplitude ranges, leading to poor correction effects from a uniform compensation method and severe measurement distortion. Second, zero-sequence current is acquired solely from a single zero-sequence transformer, lacking a verification mechanism. Transformer failures and line interference can cause zero-sequence sampling distortion, easily leading to false tripping or failure to trip of grounding protection. Third, the fixed baseline offset results in DC drift under low-load conditions. The problems are difficult to eliminate; protection action criteria mostly use fixed thresholds and do not dynamically adjust based on equipment status parameters such as contact temperature, number of mechanism operations, and ambient temperature and humidity, resulting in a significant increase in the frequency of false trips and failures to trip when equipment ages or operating conditions change; fourth, there is a lack of disconnector switch interlocking logic, so even after the disconnector switch is opened, there is no load current in the circuit, and the sampled data is still sent to the protection unit normally, which can easily lead to the problem of false triggering of protection without faults; existing technologies cannot take into account temperature drift compensation, sampling verification, adaptive criterion optimization and interlocking constraints, resulting in low current acquisition accuracy and insufficient protection fault tolerance, which cannot meet the measurement and control protection requirements for long-term stable operation of circuit breakers. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention proposes a circuit breaker current acquisition and processing method and system. It acquires raw analog current signals and multi-dimensional state parameters, preprocesses and encapsulates them into digital current data packets with state tags. Utilizing a two-dimensional correction coefficient table calibrated by temperature zone and current amplitude layers, it completes temperature drift compensation for three-phase current and zero-sequence current amplitude and phase. Through cross-verification of the three-phase synthesized zero-sequence current with the measured zero-sequence amplitude and phase using dual thresholds, it outputs effective current data. For the first time, it extracts the average value of the three-phase DC components as a baseline bias under low-load conditions, constructs a multi-source information fusion model, and adaptively optimizes the protection action criteria. Simultaneously, it integrates the disconnector switch interlocking signal, forcing the current sampling to zero and adding an interlocking mark during opening, and outputting the verified current sequence upon closing. This invention suppresses temperature drift measurement errors, achieves dual-channel zero-sequence verification to prevent sampling distortion, and dynamically optimizes the protection action criteria using multiple state parameters to reduce false trips.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for collecting and processing circuit breaker current includes:

[0006] S1: Acquire the original analog signals and multi-dimensional status parameters of the circuit breaker during operation, and output a digital current sampling data packet with status label after preprocessing;

[0007] S2: Based on the pre-stored hierarchical correction coefficients, the amplitude temperature drift compensation and phase temperature drift compensation are performed on the three-phase current and the measured zero-sequence current in the digital current sampling data package, and the corrected three-phase current sampling value and the corrected measured zero-sequence sampling value are output.

[0008] S3: Calculate the synthesized zero-sequence sampling value based on the corrected three-phase current sampling value, cross-verify the synthesized zero-sequence sampling value with the corrected measured zero-sequence sampling value, and output the effective current judgment data when the verification passes.

[0009] S4: When the preset low load condition is met for the first time, the average value of the DC component of the three-phase current in the effective current judgment data is taken as the baseline bias parameter, and the multi-dimensional state parameters are used together for fusion judgment. The influence weight of each state parameter is dynamically adjusted, and the protection action criterion is iteratively optimized. The protection action criterion is defined as triggering protection when the effective value of any phase in the corrected three-phase current sampling value exceeds 1.2 times the rated value and lasts for more than 40ms.

[0010] S5: Based on the optimized protection action criteria, collect current waveform data in real time, connect to the disconnecting switch interlocking signal, and output the current acquisition result after verification and interlocking constraints.

[0011] Specifically, the original analog signal includes analog current signals from the secondary side of the current transformer on the three-phase bushing side of the circuit breaker and analog current signals from the secondary side of the zero-sequence current transformer; the multi-dimensional state parameters include at least the circuit breaker contact temperature, the operating mechanism opening and closing coil current, the ambient temperature and humidity, and the cumulative number of circuit breaker operations.

[0012] The preprocessing specifically includes: filtering the original analog signal using an anti-aliasing filter to remove frequency components higher than 1 / 2 of the sampling frequency; synchronously sampling the filtered analog signal based on the GPS second pulse signal, with the sampling frequency set to no less than 256 points per cycle; converting the synchronously sampled signal into a digital current sampling signal using an analog-to-digital converter; performing a fast Fourier transform on the digital current sampling signal to extract the fundamental and DC components; encapsulating the extracted fundamental and DC components into a digital current sampling data packet based on a preset sampling data packet format, and attaching a status label containing the sampling time and circuit breaker identification to the digital current sampling data packet.

[0013] Specifically, the hierarchical correction coefficients are pre-stored in non-volatile memory. The hierarchical correction coefficients are calibrated hierarchically according to the ambient temperature range and the current amplitude range. The specific implementation method is as follows: In the circuit breaker factory calibration process, for multiple preset ambient temperature ranges and multiple preset current amplitude ranges, standard current signals are injected through a signal source to calculate the amplitude deviation and phase deviation between the sampled measurement value and the standard injected value. Based on the amplitude deviation and phase deviation, an amplitude correction coefficient table and a phase correction coefficient table corresponding one-to-one with each temperature range and each current amplitude range are generated. When applying the correction coefficients, the ambient temperature value is used as the first search index to determine the calibration temperature range to which the current temperature belongs, and the current amplitude value is used as the second search index to determine the preset current amplitude range to which the current current belongs. The applicable amplitude correction coefficients and phase correction coefficients at the current moment are obtained from the amplitude correction coefficient table and the phase correction coefficient table through a two-dimensional lookup table.

[0014] Specifically, the steps of S2 include:

[0015] Based on the ambient temperature value in the multidimensional state parameters at the current moment, determine the calibration temperature range to which the current temperature belongs, call the amplitude correction coefficient under the calibration temperature range, and multiply the amplitudes of the three-phase current and the measured zero-sequence current in the digital current sampling data packet by the corresponding amplitude correction coefficients respectively to obtain the amplitude-compensated three-phase current amplitude and the measured zero-sequence current amplitude.

[0016] Based on the ambient temperature value in the multidimensional state parameters at the current moment, determine the calibration temperature range to which the current temperature belongs, call the phase correction coefficient under the calibration temperature range, and superimpose the corresponding phase correction coefficients on the phases of the three-phase current and the measured zero-sequence current in the digital current sampling data packet to obtain the phase-compensated three-phase current phase and the measured zero-sequence current phase.

[0017] The amplitude of the three-phase current after amplitude compensation is combined with the phase of the three-phase current after phase compensation to reconstruct the corrected three-phase current sample value.

[0018] The measured zero-sequence current amplitude after amplitude compensation is combined with the measured zero-sequence current phase after phase compensation to reconstruct the corrected measured zero-sequence sampling value.

[0019] Specifically, the steps of S3 include:

[0020] The A-phase current sample value, B-phase current sample value, and C-phase current sample value of the corrected three-phase current sample values ​​at the same sampling time are summed by phasor operation to obtain the composite zero-sequence phasor value at that sampling time.

[0021] Calculate the relative amplitude deviation and phase deviation angle between the synthesized zero-sequence sampled value and the corrected measured zero-sequence sampled value. When the relative amplitude deviation is less than a preset first threshold and the phase deviation angle is less than a preset second threshold, the measured zero-sequence sampled value at the current moment is determined to be reliable and the verification is passed.

[0022] After the verification is passed, the corrected three-phase current sampling value and the corrected measured zero-sequence sampling value are used together to form the effective current determination data at the current sampling time and output it.

[0023] Specifically, the preset low load condition is defined as follows: the circuit breaker is in the closed operation state and the effective value of each phase of the corrected three-phase current sampling value is less than 10% of the rated current and the duration is greater than or equal to 5s; when the low load condition is met for the first time, the arithmetic mean of the DC component sampling points of the three-phase current in the effective current judgment data within one power frequency cycle time window before and after the current time is used as the A-phase baseline offset parameter, B-phase baseline offset parameter and C-phase baseline offset parameter.

[0024] Specifically, the joint multi-dimensional state parameters are used for fusion discrimination, the influence weight of each state parameter is dynamically adjusted, and the protection action criteria are iteratively optimized, including:

[0025] A comprehensive operational characteristic judgment calculation link is constructed, with circuit breaker contact temperature, characteristic peak value of operating mechanism opening and closing coil current, ambient temperature and humidity, and cumulative number of circuit breaker operations as input characteristic parameters, and action sensitivity coefficient of protection action criterion as output target.

[0026] A feature dimensionality reduction method based on principal component analysis algorithm is used to compress the relevant features in multidimensional state parameters into mutually independent principal component features.

[0027] The initial weight coefficients of each principal component feature in the fusion discrimination are calculated based on the entropy weight method.

[0028] After each protection action event occurs, the actual judgment result of the protection action criterion and the actual fault occurrence are recorded. When the frequency of false operation increases, the weight coefficients of each principal component feature quantity are backpropagated and corrected based on the multidimensional state parameter feature values ​​at the time of the false operation event.

[0029] Specifically, the iterative optimization triggering condition in S4 is as follows: when the cumulative number of protection actions reaches a preset batch action number threshold or the cumulative false trip frequency exceeds a preset false trip frequency threshold, a weight coefficient iterative update process is triggered; after each iterative update, the updated weight coefficient is substituted into the operation feature comprehensive judgment calculation link, the action sensitivity coefficient of the protection action criterion is recalculated, and the multiple threshold of rated current in the protection action criterion is adaptively corrected according to the updated action sensitivity coefficient; the iterative update process is repeated until the false trip frequency is less than or equal to 1 in 100 consecutive protection operations, the current weight coefficient and action sensitivity coefficient are frozen, and the iterative optimization is stopped; the false trip frequency is the number of times the protection is falsely triggered when the criterion condition is not met or when the condition is met but the protection fails to operate.

[0030] Specifically, the steps of S5 include:

[0031] Under normal operating conditions of the circuit breaker, the current signal of the secondary side of the current transformer of the three-phase bushing of the circuit breaker is collected, a real-time current waveform sampling data stream is generated, and the verified real-time effective current sampling sequence is obtained.

[0032] The auxiliary contact status signal of the disconnecting switch is acquired in real time through the input interface of the circuit breaker controller; the auxiliary contact status signal includes a first status signal indicating that the disconnecting switch is in the closed position and a second status signal indicating that the disconnecting switch is in the open position.

[0033] The acquired auxiliary contact status signal is judged. When the auxiliary contact status signal is judged to be the second status signal, a valid flag for the disconnecting switch interlock is generated and the valid flag is set to the valid state. When the auxiliary contact status signal is judged to be the first status signal, the valid flag for the disconnecting switch interlock is kept in the unset state.

[0034] When the read disconnector interlocking effective flag is not set, the real-time effective current sampling sequence is directly output to the circuit breaker protection control unit as the effective current acquisition result.

[0035] When the read disconnector interlocking effective flag is in the set state, all sampled values ​​in the real-time effective current sampling sequence at the current moment are forcibly modified to zero, generating a forced zero-set current sampling sequence carrying the interlocking action flag, and then the forced zero-set current sampling sequence is output to the circuit breaker protection control unit as an effective current acquisition result.

[0036] A circuit breaker current acquisition and processing system, comprising:

[0037] The signal acquisition module is used to acquire the original analog current signal and multi-dimensional state parameters of the circuit breaker;

[0038] The layered temperature drift correction module uses real-time ambient temperature and current amplitude as a two-dimensional search index to match the corresponding correction coefficients. It performs amplitude multiplication coefficient compensation and phase superposition coefficient compensation on the three-phase current and the measured zero-sequence current respectively, and then reconstructs the amplitude and phase information, outputting the three-phase current sampled value and the measured zero-sequence sampled value after temperature drift correction.

[0039] The zero-sequence mutual verification module is used to sum the phasors of the corrected three-phase currents at the same time to obtain the composite zero-sequence sampled value, calculate the relative amplitude deviation and phase deviation angle between the composite zero-sequence sampled value and the measured zero-sequence sampled value, and generate effective current judgment data output after the verification is qualified.

[0040] The criterion optimization module is used to extract the average value of the DC component of the three-phase current as the baseline bias parameter, build a comprehensive judgment calculation link for operation characteristics with various equipment status parameters as input, and iteratively correct the protection action sensitivity coefficient and protection action criterion based on the cumulative protection action and the frequency of false action triggering weights.

[0041] The interlocking constraint output module collects real-time current waveforms to generate effective current sampling sequences, reads the interlocking signals of the opening and closing status of the disconnecting switch auxiliary contacts, directly outputs the current sampling sequence to the protection unit when the disconnecting switch is closed, and forces all sampling values ​​to zero and adds interlocking action markers before outputting when the disconnecting switch is opened.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] 1. This invention proposes a circuit breaker current acquisition and processing method. This invention effectively improves the measurement accuracy and data reliability of circuit breaker current acquisition. By calibrating a two-dimensional correction coefficient table according to temperature and current amplitude, it synchronously completes amplitude and phase temperature drift compensation for three-phase and zero-sequence currents, solving the measurement deviation caused by transformer temperature drift. It employs GPS-synchronized high-frequency sampling combined with anti-aliasing filtering, and uses FFT to separate the fundamental frequency and DC component. The three-phase DC average value is extracted under low-load conditions as a baseline bias to suppress DC drift interference. Simultaneously, a dual-threshold cross-verification mechanism for synthesized zero-sequence and measured zero-sequence amplitude and phase is set up, which can quickly identify sampling distortion caused by transformer abnormalities and electromagnetic interference. This ensures the authenticity and validity of current sampling data throughout the entire process of signal acquisition, error compensation, and data verification.

[0044] 2. This invention proposes a circuit breaker current acquisition and processing method. This invention can adaptively optimize protection action criteria, reduce the probability of protection maloperation and failure to operate, and add interlocking logic to improve the adaptability of operating conditions. It constructs a multi-source discrimination model that integrates contact temperature, mechanism coil current, ambient temperature and humidity, and number of operations. After PCA dimensionality reduction and entropy weighting, the feature weights are iteratively corrected according to the protection action and maloperation frequency, and the current protection threshold is adaptively corrected. This overcomes the shortcomings of traditional fixed criteria that cannot adapt to equipment aging and operating condition fluctuations. In addition, it integrates the disconnecting switch position interlocking signal, forces the current sampling to zero and adds an interlocking mark when the switch is opened, avoids protection mal-triggering under no-load conditions, and improves the overall reliability of circuit breaker current measurement and control and fault protection. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of a circuit breaker current acquisition and processing method according to the present invention;

[0046] Figure 2 This is a flowchart illustrating the principle of a circuit breaker current acquisition and processing method according to the present invention.

[0047] Figure 3 This is a schematic diagram of a circuit breaker current acquisition and processing system according to the present invention. Detailed Implementation

[0048] Example 1

[0049] Please see Figures 1-2 The present invention provides an embodiment of a circuit breaker current acquisition and processing method, comprising the following steps:

[0050] S1: Acquire the original analog signals and multi-dimensional status parameters of the circuit breaker during operation, and output a digital current sampling data packet with status label after preprocessing;

[0051] The original analog signal includes analog current signals from the secondary side of the current transformer on the three-phase bushing side of the circuit breaker and analog current signals from the secondary side of the zero-sequence current transformer; the multi-dimensional state parameters include at least the circuit breaker contact temperature, the operating mechanism opening and closing coil current, the ambient temperature and humidity, and the cumulative number of circuit breaker operations.

[0052] The preprocessing specifically includes: filtering the original analog signal using an anti-aliasing filter to remove frequency components higher than 1 / 2 of the sampling frequency; synchronously sampling the filtered analog signal based on the GPS second pulse signal, with the sampling frequency set to no less than 256 points per cycle; converting the synchronously sampled signal into a digital current sampling signal using an analog-to-digital converter; performing a fast Fourier transform on the digital current sampling signal to extract the fundamental and DC components; encapsulating the extracted fundamental and DC components into a digital current sampling data packet based on a preset sampling data packet format, and attaching a status tag containing the sampling time and circuit breaker identification to the digital current sampling data packet. The analog-to-digital converter and fast Fourier transform are existing technologies in this field and are not inventive solutions of this application, and will not be elaborated upon here.

[0053] This embodiment uses a 10kV indoor vacuum circuit breaker as the actual application carrier to carry out a complete process description. The circuit breaker is equipped with A, B, and C three-phase bushing current transformers, an independent zero-sequence current transformer, a platinum resistance temperature sensor built into the contacts, a current sampling circuit for the opening and closing coils, a cabinet temperature and humidity acquisition module, and a mechanical operation counting unit. All acquisition circuits are uniformly connected to the circuit breaker's built-in edge acquisition controller. The entire set of acquisition hardware is integrated and installed inside the circuit breaker's secondary control compartment, avoiding electromagnetic interference and line loss problems caused by external wiring on site. Under normal energized closing operation of the circuit breaker, the primary winding of the three-phase bushing side current transformer is connected in series to the main circuit. After the load current of the main circuit passes through the transformer core, it outputs a continuous analog current signal on the secondary side according to a fixed transformation ratio. The zero-sequence current transformer is mounted on the outside of the three-phase outgoing cable and induces a zero-sequence analog current signal through the three-phase current vector difference. The analog current signals output by both types of transformers are connected to the analog input terminal of the edge acquisition controller through shielded twisted pair cables. The outer metal shielding layer of the shielded cable is grounded at one end throughout to suppress high-frequency electromagnetic interference generated by the opening and closing of switches and the switching of contactors inside the high-voltage cabinet, ensuring that there is no noise superposition during the acquisition of the original analog signal.

[0054] For multi-dimensional state parameters, high-temperature platinum resistance temperature sensors are pre-embedded inside the moving and stationary contacts of the circuit breaker. The sensor leads are led out through the reserved wiring channel in the vacuum interrupter to the temperature measurement channel of the acquisition controller to collect the contact operating temperature value in real time. A high-precision sampling resistor is connected in parallel across the opening and closing coils of the operating mechanism. The controller calculates the real-time current waveform of the opening and closing coils by converting the voltage across the sampling resistors, and records the current change curve of the coil throughout the entire energization process. Temperature and humidity acquisition probes are installed at two locations inside the cabinet, one at a high position and one at a low position. The high probe collects the temperature and humidity of the heat dissipation area at the top of the cabinet, and the low probe collects the temperature and humidity of the water and dust accumulation area at the bottom of the mechanism. The average value of the two collection values ​​is taken as the overall ambient temperature and humidity parameter. A mechanical counting contact is installed at the main shaft of the circuit breaker operating mechanism. Each time a complete opening and closing operation is completed, the contact closes once. The controller's built-in register accumulates and records the total number of contact closures, generating the cumulative operation count parameter of the circuit breaker. All multi-dimensional state parameters and the original analog current signal are collected in a completely synchronized timing sequence, with no time acquisition offset.

[0055] Furthermore, when using an anti-aliasing filter to filter the original analog signal and remove frequency components higher than half the sampling frequency, this embodiment sets the number of sampling points per cycle to 256, the power frequency standard to 50Hz, and the single cycle time length to 20ms, corresponding to an overall sampling frequency of 12800Hz. Half the sampling frequency is 6400Hz. Therefore, the anti-aliasing filter is designed as an eighth-order active low-pass anti-aliasing filter circuit with a cutoff frequency of 6400Hz. The four independent filter channels correspond to the three-phase current analog signals A, B, and C and the zero-sequence current analog signal, respectively. The hardware parameters of the four filter circuits are completely identical, eliminating the inherent amplitude and phase deviations between channels. After the original analog current signal is input to the filter circuit, it first passes through a first-order RC buffer circuit to complete impedance matching. After matching, the signal is sent to a filter link composed of multi-stage operational amplifiers. The filter link attenuates high-frequency noise above 6400Hz step by step, including the arc interference of several kilohertz generated by cabinet switch operation, and the interference from surrounding frequency converters and motors. All components of the equipment, including high-frequency harmonics and line-coupled radio frequency interference, that are higher than the cutoff frequency are attenuated by the filtering circuit to a negligible amplitude range. After filtering, the controller reads the continuous analog waveform of the filter output port in real time and continuously monitors the smoothness of the filtered signal waveform. In this embodiment, the glitch amplitude threshold is set to 0.5% of the secondary rated current of the transformer, which is converted to 0.005 times the full scale of the acquisition chip corresponding to the analog voltage sampling level. The judgment logic is as follows: if the instantaneous glitch amplitude of any sampling point within a single power frequency cycle continuously exceeds 0.5%, it is determined that there is abnormal high-frequency interference. If the high-frequency glitch amplitude exceeds 0.5% for more than ten consecutive power frequency cycles, the controller automatically generates a filtering abnormality alarm mark and synchronously writes it into the status label field of the digital current sampling data packet, which is convenient for the back-end maintenance personnel to identify transformer, line, or filter hardware faults. The smoothed analog current signal output by the filter is directly sent to the back-end synchronous sampling analog-to-digital conversion channel without going through intermediate storage buffer, thus avoiding signal delay and distortion.

[0056] Furthermore, the step of converting the synchronously sampled signals into digital current sampling signals via an analog-to-digital converter includes: In this embodiment, a four-channel independent synchronous sampling analog-to-digital converter chip is selected. The four conversion channels share the same set of sampling trigger signals controlled by GPS second pulse synchronization. After filtering, the four analog current signals are respectively connected to the corresponding input channels of the analog-to-digital converter. Overvoltage and overcurrent protection diodes are added to the input ports of each channel to prevent damage to the conversion chip from short circuits and lightning surges. When the synchronous sampling time slot trigger signal arrives, the analog-to-digital converter simultaneously latches the current analog voltage values ​​of the four channels. The successive approximation conversion circuit converts continuous analog quantities into binary digital quantities. After the conversion is completed, the corresponding channel digital sample value is cached in the register. The entire conversion time is controlled within 10% of a single sampling time slot and will not occupy the next sampling trigger timing. Each digital sample value is accompanied by an independent channel identifier to distinguish four types of sample data: A phase, B phase, C phase, and zero sequence. After the conversion is completed, all digital sample signals are uniformly sent to the floating-point memory block of the controller to wait for fast Fourier transform operation. During the conversion process, the operating temperature of the analog-to-digital converter chip itself is recorded and stored in the multi-dimensional state parameter set.

[0057] Furthermore, the detailed implementation process of performing Fast Fourier Transform (FFT) calculation on the digital current sampling signal to extract the fundamental and DC components includes: A circular data buffer of 256 points is opened in the controller memory, continuously storing four synchronous digital current sampling signals of a complete power frequency cycle. When the buffer is filled with all 256 sampling data points, the FFT operation subroutine is automatically triggered. The operation subroutine first preprocesses the 256-point time-domain digital sampling signal to eliminate calculation errors caused by the zero-point offset of the sampling start. Then, it performs an iterative FFT operation from the time domain to the frequency domain, decomposing the single-cycle time-domain current waveform into a set of frequency-domain data with different frequency components. In the frequency-domain data, 0Hz corresponds to the DC component, 50Hz corresponds to the power frequency fundamental component, and the remaining frequencies correspond to the 2nd, 3rd, 5th, and other higher harmonic components. After the calculation is completed, the subroutine extracts the amplitude and phase offset of the 0Hz DC component separately, and extracts the amplitude and phase parameters of the 50Hz fundamental component. The higher harmonic components are temporarily cached and used only for fault harmonic analysis. All extracted DC and fundamental components carry the corresponding channel identifier and are bound to the current cycle GPS precise timestamp for storage to prevent confusion of multi-cycle and multi-channel data. If a large number of missing or abrupt changes in single-cycle sampling data are detected during the calculation, the subroutine marks the fundamental and DC components of that cycle as invalid values.

[0058] In this embodiment, a unified standard sampling data packet encapsulation format is pre-embedded in the controller's non-volatile memory chip. The data packet is divided into four fixed partitions: header field, data body field, status label field, and checksum field. The length and data type of each partition field are fixed to ensure consistent parsing during local storage and background transmission. The header field stores the data packet length, sampling cycle number, and number of sampling channels. The data body field sequentially writes the fundamental amplitude, fundamental phase, and DC component amplitude corresponding to the four currents of phase A, phase B, phase C, and zero sequence. All values ​​are uniformly stored in floating-point data type, retaining a preset decimal precision. The status label field is further subdivided into three subfields. The first subfield is the sampling time subfield, which is filled with GPS second pulse synchronization. The generated data includes complete time information accurate to microseconds, including year, month, day, hour, minute, second, and microsecond. The second category is the circuit breaker identification subfield, which stores the factory-fixed unique equipment serial number, switch cabinet number, and primary circuit bay number of the circuit breaker, used to distinguish the equipment source when multiple circuit breakers are collected concurrently. The third category is the multi-source status flag subfield, which synchronously writes the original values ​​of all multi-dimensional status parameters collected in the current cycle, including real-time contact temperature, peak current of opening and closing coils, ambient temperature and humidity, and cumulative number of operations, while superimposing various fault flags such as the aforementioned filtering anomaly and invalid sampling data. The check tail field uses a cyclic redundancy check algorithm to check and calculate the entire contents of the data packet, generating a fixed-length check code to be appended to the end of the data packet.

[0059] S2: Based on the pre-stored hierarchical correction coefficients, the amplitude temperature drift compensation and phase temperature drift compensation are performed on the three-phase current and the measured zero-sequence current in the digital current sampling data package, and the corrected three-phase current sampling value and the corrected measured zero-sequence sampling value are output.

[0060] The tiered correction coefficients are pre-stored in non-volatile memory. The tiered correction coefficients are calibrated in tiers based on the ambient temperature range and the current amplitude range. Specifically, the implementation method is as follows:

[0061] A1: During the factory calibration of the circuit breaker, for multiple preset ambient temperature ranges and multiple preset current amplitude ranges, a standard current signal is injected through a signal source to calculate the amplitude deviation and phase deviation between the sampled measurement value and the standard injected value. Based on the amplitude deviation and phase deviation, an amplitude correction coefficient table and a phase correction coefficient table corresponding one-to-one with each temperature range and each current amplitude range are generated.

[0062] Furthermore, in this embodiment, a constant temperature high and low temperature test chamber is built in the factory calibration laboratory for the 10kV vacuum circuit breaker. The internal temperature of the test chamber can be continuously adjusted from -40℃ to 70℃. It is equipped with a high-precision standard three-phase current source, a standard zero-sequence current signal source, and a high-precision reference sampling and measurement device. The metrological accuracy level of the reference device is two levels higher than that of the circuit breaker acquisition controller, and it serves as a reference for the true value of the calibration standard.

[0063] Furthermore, the ambient temperature calibration range is pre-defined, and eight continuous non-overlapping temperature ranges are divided according to the actual operating conditions of the equipment, namely -40℃~-20℃, -20℃~0℃, 0℃~15℃, 15℃~30℃, 30℃~45℃, 45℃~55℃, 55℃~65℃, and 65℃~70℃, which fully covers all actual operating temperature scenarios such as low temperature outdoor in the north, high temperature indoor in the south, high temperature in the sealed switchgear in summer, and low temperature in the basement in winter.

[0064] Furthermore, the current amplitude calibration range is divided into nine levels: the range is divided according to the rated current of the circuit breaker from 0 to 1.5 times the rated current, including 0~10% rated current, 10%~25% rated current, 25%~40% rated current, 40%~60% rated current, 60%~80% rated current, 80%~100% rated current, 100%~120% rated current, 120%~140% rated current, and 140%~150% rated current, taking into account all operating conditions such as low load no-load operation, normal rated load, and overload protection range.

[0065] Furthermore, during calibration, the complete circuit breaker is placed inside a high and low temperature test chamber. After sealing the chamber, the first temperature range is set and kept at a constant temperature for two hours to allow the overall temperature of the circuit breaker's current transformers, acquisition circuit boards, and analog-to-digital conversion chips to be completely consistent with the ambient temperature inside the chamber, thus eliminating calibration errors caused by temperature gradients in the components. After the temperature stabilizes, a standard current source sequentially injects standard amplitude and standard phase power frequency current signals of each range into the secondary side of the three current transformers A, B, and C, and the secondary side of the zero-sequence current transformer. The current amplitude of each range remains stable for thirty seconds, and the reference measurement device simultaneously records the true value of the standard injected current. The circuit breaker's own acquisition controller simultaneously executes the S1 acquisition preprocessing process to generate the corresponding digital current sampling data packet.

[0066] Furthermore, the controller automatically compares the amplitude and phase values ​​sampled within the data packet with the true value of the standard injected current recorded by the reference device, calculating the amplitude deviation and phase deviation values ​​for the current temperature range and the current current amplitude range. Based on the deviation values, a compensation correction coefficient is calculated in reverse. If the sampled amplitude is less than the true value of the standard injected current, the amplitude correction coefficient is greater than 1, amplifying the sampled amplitude to match the standard value; if the sampled amplitude is greater than the true value of the standard injected current, the amplitude correction coefficient is less than 1, reducing the sampled amplitude. If the phase deviation is lagging, the correction coefficient is positive, and the sampled phase is superimposed for positive compensation; if the phase is leading, the correction coefficient is negative, offsetting the leading offset. For each temperature range... For each current amplitude range, independent amplitude correction coefficients and phase correction coefficients for phases A, B, C, and zero sequence are generated. All coefficients are stored in a two-dimensional table structure, with row indices corresponding to temperature range numbers and column indices corresponding to current amplitude range numbers, forming a complete two-dimensional table of amplitude correction coefficients and a two-dimensional table of phase correction coefficients. After all eight temperature ranges and nine current amplitude ranges are calibrated, the two two-dimensional coefficient tables are completely written into the circuit breaker's built-in non-volatile flash memory chip. The data is permanently saved after the flash memory chip is powered off, eliminating the need for recalibration each time it is powered on. At the same time, the temperature range boundary values ​​and current amplitude range boundary values ​​are stored synchronously as judgment thresholds for table lookup during on-site operation.

[0067] A2: When applying the correction coefficient, the ambient temperature value is used as the first search index to determine the calibration temperature range to which the current temperature belongs, and the current amplitude value is used as the second search index to determine the preset current amplitude range to which the current current belongs. The applicable amplitude correction coefficient and phase correction coefficient at the current moment are obtained from the amplitude correction coefficient table and the phase correction coefficient table through a two-dimensional lookup table.

[0068] Furthermore, the step of determining the calibrated temperature range to which the current temperature belongs using the ambient temperature value as the first retrieval index includes: after the controller completes the encapsulation of each frame of digital current sampling data packet, it synchronously reads the real-time ambient temperature and humidity parameters stored in the status label of the data packet, extracts the average ambient temperature as the original retrieval value, and compares and matches this temperature value with the upper and lower boundary thresholds of the eight temperature ranges one by one. The matching logic judges in order from low temperature to high temperature. When the real-time ambient temperature value is greater than the lower limit of the range and less than or equal to the upper limit of the range, it is determined that the current operating condition falls into the temperature range of that range, and the index number of the two-dimensional coefficient table row corresponding to the temperature range is locked. If the real-time temperature is exactly equal to the boundary threshold of the range, it is automatically assigned to the higher temperature range for retrieval. If the equipment operating temperature exceeds the factory-calibrated complete range of -40℃ to 70℃, the controller determines that there is no matching correction coefficient for the over-temperature, generates a temperature drift correction failure mark and writes it into the status label of the data packet, and the amplitude and phase compensation links directly use the uncorrected original sampling value. At the same time, the over-temperature alarm is uploaded to the background monitoring system. After the matching is completed, the temperature range index number is temporarily stored in the memory register.

[0069] Furthermore, the step of determining the preset current amplitude range to which the current belongs using the current amplitude as the second retrieval index includes: the controller extracts the uncorrected fundamental amplitudes of the four current channels A, B, C, and zero-sequence from the main data field of the data packet; the four channels independently execute the range retrieval logic without interfering with each other; the single-channel amplitude retrieval logic compares the current channel's fundamental amplitude with the nine-level current amplitude range boundary thresholds sequentially; if the amplitude is greater than the lower limit of the range and less than or equal to the upper limit of the range, the corresponding column index number is locked; if the current amplitude exceeds 150% of the rated current, it is determined to be out of calibration range, the channel is marked as having failed correction, and only the original amplitude phase is output; if the current amplitude is less than 0% of the rated current, i.e., a negative value offset is sampled, and the channel is uniformly classified into the lowest 0~10% range for correction.

[0070] Furthermore, each of the four channels generates an independent current amplitude range column index. The same temperature row index is combined with the four independent column indexes to realize differentiated two-dimensional lookup tables for the four currents. This adapts to the precise correction of different current amplitudes under various operating conditions such as three-phase load imbalance and zero-sequence leakage, and avoids the problem of coarse compensation where a large current in one phase and a small current in two phases share the same set of correction coefficients. All the index numbers of the four channels are temporarily cached and used for lookup tables for the four currents respectively.

[0071] The specific steps of S2 include:

[0072] S2.1: Determine the calibration temperature range to which the current temperature belongs based on the ambient temperature value in the multidimensional state parameters at the current moment, call the amplitude correction coefficient under the calibration temperature range, and multiply the amplitudes of the three-phase current and the measured zero-sequence current in the digital current sampling data packet by the corresponding amplitude correction coefficients respectively to obtain the amplitude of the three-phase current and the measured zero-sequence current after amplitude compensation.

[0073] Furthermore, the specific steps of S2.1 include:

[0074] (1) Read the real-time ambient temperature within the current data packet status label, complete the temperature range matching, and obtain the global unified temperature row index;

[0075] (2) Read the original fundamental amplitude values ​​of the four channels A, B, C and zero sequence respectively, and match the current amplitude range of the four channels independently to obtain the independent column index of each channel;

[0076] (3) Using the temperature row index and the A-phase current column index as two-dimensional search coordinates, access the two-dimensional table of amplitude correction coefficients and extract the A-phase exclusive amplitude correction coefficients; similarly, look up the table in turn to extract the amplitude correction coefficients corresponding to the B-phase, C-phase and zero-sequence current respectively. The four correction coefficients are stored separately and are not mixed with each other.

[0077] (4) Perform amplitude compensation operation on the original fundamental amplitude of the four channels respectively, multiply the original fundamental amplitude of phase A in the data packet by the amplitude correction coefficient of phase A to obtain the amplitude of phase A after amplitude compensation; among them, phases B, C and zero sequence use the same operation logic to obtain their respective amplitude of phase A after amplitude compensation.

[0078] (5) Temporarily buffer the amplitude values ​​after the four-channel amplitude compensation is completed, and simultaneously retain the original uncompensated amplitude data for fault tracing, and wait for the phase compensation to be completed before reconstructing the sampled values;

[0079] (6) If any channel is determined to have no effective correction coefficient in the over-temperature or over-current range during the table lookup process, the amplitude compensation value of that channel is directly assigned to the original amplitude value, and a compensation failure mark is marked on the corresponding channel.

[0080] S2.2: Determine the calibration temperature range to which the current temperature belongs based on the ambient temperature value in the multidimensional state parameters at the current moment, call the phase correction coefficient under the calibration temperature range, and superimpose the corresponding phase correction coefficients on the phases of the three-phase current and the measured zero-sequence current in the digital current sampling data packet to obtain the phase-compensated three-phase current phase and the measured zero-sequence current phase.

[0081] Furthermore, the specific steps of S2.2 include:

[0082] (1) Reuse the global temperature row index output by S2.1. The global temperature row index is the interval number generated after matching eight preset temperature ranges based on the real-time ambient temperature. The number range is 1 to 8, which correspond to eight calibrated temperature ranges: -40℃ to -20℃, -20℃ to 0℃, 0℃ to 15℃, 15℃ to 30℃, 30℃ to 45℃, 45℃ to 55℃, 55℃ to 65℃, and 65℃ to 70℃. The current temperature range can be locked by directly calling the global temperature row index, which reduces the memory and timing resources occupied by the controller for repeated logic operations.

[0083] (2) The current amplitude column indexes of the four channels output by S2.1 are independently matched. The four channels correspond to the A-phase current channel, B-phase current channel, C-phase current channel and zero-sequence current channel respectively. The current amplitude column index of each channel is the interval number generated after matching the original fundamental amplitude of the corresponding channel with the nine preset current amplitude intervals. The number range is 1 to 9, which correspond to the nine calibrated current amplitude intervals of 0 to 10% rated current, 10% to 25% rated current, 25% to 40% rated current, 40% to 60% rated current, 60% to 80% rated current, 80% to 100% rated current, 100% to 120% rated current, 120% to 140% rated current and 140% to 150% rated current respectively. There is no need to extract the original fundamental amplitude of each channel in the digital current sampling data package again and repeat the current amplitude interval comparison and judgment.

[0084] (3) Using the global temperature row index as the row positioning parameter for the two-dimensional lookup table and the A-phase current amplitude column index as the column positioning parameter for the two-dimensional lookup table, a two-dimensional retrieval coordinate access phase correction coefficient two-dimensional table is formed to extract the A-phase exclusive phase correction coefficient; using the global temperature row index in combination with the B-phase current amplitude column index to access the phase correction coefficient two-dimensional table to extract the B-phase exclusive phase correction coefficient; using the global temperature row index in combination with the C-phase current amplitude column index to access the phase correction coefficient two-dimensional table to extract the C-phase exclusive phase correction coefficient; using the global temperature row index in combination with the zero-sequence current amplitude column index to access the phase correction coefficient two-dimensional table to extract the zero-sequence exclusive phase correction coefficient. All four phase correction coefficients have positive and negative offset indicators. The phase offset is uniformly measured in degrees. The phase correction coefficient values ​​are stored with two decimal places.

[0085] (4) Read the original fundamental phase value of phase A stored in the digital current sampling data packet, and perform an addition operation on the original fundamental phase value of phase A and the extracted phase A correction coefficient. The result is the phase A value after phase compensation. Read the original fundamental phase value of phase B stored in the digital current sampling data packet, and perform an addition operation on the original fundamental phase value of phase B and the extracted phase B correction coefficient. The result is the phase B value after phase compensation. Read the original fundamental phase value of phase C stored in the digital current sampling data packet, and perform an addition operation on the original fundamental phase value of phase C and the extracted phase C correction coefficient. The result is the phase C value after phase compensation. Read the original fundamental phase value of zero sequence stored in the digital current sampling data packet, and perform an addition operation on the original fundamental phase value of zero sequence and the extracted zero sequence phase correction coefficient. The result is the zero sequence phase value after phase compensation. The four channels independently complete the phase superposition operation, and the operation timing is executed synchronously.

[0086] (5) The phase values ​​of the four channels after compensation are calculated and superimposed are uniformly subjected to boundary normalization period reduction processing. The period reduction processing is fixed with 360° as the complete power frequency phase period. When the phase value of a single channel after compensation is greater than 360°, the phase value is continuously subtracted by 360° until the processed phase value falls within the range of 0° to 360°. When the phase value of a single channel after compensation is less than 0°, the phase value is continuously increased by 360° until the processed phase value falls within the range of 0° to 360°. The phase values ​​of the four channels after normalization are uniformly limited to the standard phase range of 0° to 360°. The storage accuracy of the phase values ​​after normalization is consistent with the original fundamental phase.

[0087] (6) Store the compensation phase values ​​of the four channels that have completed the normalization operation into the controller’s dedicated temporary memory cache area, synchronously read the channel compensation failure flag generated by the flag, and bind the compensation failure flag of the corresponding channel with the normalized compensation phase value for storage.

[0088] S2.3: Combine and reconstruct the amplitude of the three-phase current after amplitude compensation with the phase of the three-phase current after phase compensation, and output the corrected three-phase current sample value; execute independently for phases A, B, and C in sequence. The reconstruction logic is fully explained using phase A as an example. The process for phases B and C is completely the same.

[0089] Furthermore, the specific steps in S2.3 include:

[0090] (1) Retrieve the amplitude of phase A after amplitude compensation, and at the same time retrieve the normalized phase of phase A after compensation stored in the same temporary memory cache area. The amplitude of phase A after amplitude compensation is the fundamental amplitude of the secondary side of the transformer after the amplitude correction coefficient multiplication process, and the normalized phase after compensation is the phase value after 360° periodic normalization and the value range is fixed from 0° to 360°.

[0091] (2) Call the phasor reconstruction subroutine built into the controller. This phasor reconstruction subroutine is a factory-fixed calculation program. The program has no external adjustable configuration parameters. It only receives two sets of input parameters: the amplitude of phase A after amplitude compensation and the phase after normalization compensation corresponding to a single channel. The program combines the amplitude data and phase data of a single channel into a complete phasor sample value that can be used for vector calculation according to the fixed calculation logic. The phasor sample value simultaneously encapsulates two core dimension information: amplitude and phase offset. The generated phasor sample value can be directly used as the standard input data for three-phase vector summation calculation to calculate the synthesized zero-sequence phasor value.

[0092] (3) After the phasor reconstruction subroutine is completed, output the complete phase current sampling value after phase A correction, read the channel compensation effective or compensation failure flag generated synchronously and bound to the phase A channel data during the operation, attach the flag to the internal identifier field of the complete phase current sampling value after phase A correction, and complete the encapsulation output of the single-channel phase A current sampling value.

[0093] (4) Repeatedly execute the entire operation process of retrieving cache parameters, calling phasor reconstruction subroutine, generating the corresponding channel corrected phase current sampling value and attaching channel compensation flag, and perform phase current reconstruction operation on phase B and phase C in sequence to generate the corrected complete phase current sampling value of phase B and the corrected complete phase current sampling value of phase C respectively, and obtain three independent corrected three-phase current sampling values. The three sampling values ​​are each bound to the compensation status flag of the corresponding channel.

[0094] (5) Read the unified sampling time timestamp stored in the header of the current frame digital current sampling data packet. The timestamp is generated by GPS second pulse synchronization with a precision of microseconds. Store the three corrected three-phase current sampling values ​​into the dedicated corrected current buffer block divided by the controller. Pack and bind the unified sampling time timestamp with the three corrected three-phase current sampling values ​​for storage, so that the three three-phase current sampling values ​​share the same set of timestamp identifiers, ensuring that the three-phase data timing is fully aligned.

[0095] S2.4: Combine and reconstruct the measured zero-sequence current amplitude after amplitude compensation and the measured zero-sequence current phase after phase compensation, and output the corrected measured zero-sequence sampling value.

[0096] S3: Calculate the synthesized zero-sequence sampling value based on the corrected three-phase current sampling value, cross-verify the synthesized zero-sequence sampling value with the corrected measured zero-sequence sampling value, and output the effective current judgment data when the verification passes.

[0097] The specific steps of S3 include:

[0098] S3.1: Perform phasor summation on the A-phase current sample value, B-phase current sample value, and C-phase current sample value of the corrected three-phase current sample values ​​at the same sampling time to obtain the composite zero-sequence phasor value at that sampling time. In this embodiment, it is strictly limited that the A, B, and C phase current sample values ​​after correction must come from the same GPS synchronous sampling cycle and have completely consistent timestamps to prevent the composite zero-sequence calculation from being distorted due to the misalignment of three-phase data across cycles and time slots.

[0099] Furthermore, the specific steps of S3.1 include:

[0100] (1) Read the three sets of complete phasor sample values ​​of phase A, phase B and phase C with complete timing alignment from the corrected current buffer block. Each set of phasors includes the compensated amplitude and normalized 0~360° phase.

[0101] (2) According to the unified 50Hz three-phase standard phase reference, the complete phasor sampling values ​​of phases A, B and C are split and converted respectively. The reference parameters are fixed as 0° for phase A reference phase, 120° for phase B reference phase, and 240° for phase C reference phase. After the split and conversion, each set of phasor sampling values ​​will correspond to a set of real components on the horizontal axis and imaginary components on the vertical axis of a plane rectangular coordinate system, respectively obtaining the real component of phase A, the imaginary component of phase A, the real component of phase B, the imaginary component of phase B, the real component of phase C, and the imaginary component of phase C.

[0102] (3) Add all the real components of phase A, phase B, and phase C together to obtain the total real component of the composite zero sequence; add all the imaginary components of phase A, phase B, and phase C together to obtain the total imaginary component of the composite zero sequence.

[0103] (4) Combine the calculated total real component of the synthesized zero sequence with the total imaginary component of the synthesized zero sequence to obtain the complete synthesized zero sequence phasor. At the same time, calculate the amplitude and phase values ​​corresponding to the synthesized zero sequence sample value. The amplitude and phase values ​​are the synthesized zero sequence phasor values ​​formed after the three-phase current vectors cancel each other out in theory.

[0104] (5) Store the calculated composite zero-sequence amplitude and composite zero-sequence phase into the controller's temporary memory area, retrieve the generated and saved corrected measured zero-sequence sampled values, and store the composite zero-sequence amplitude, composite zero-sequence phase and corrected measured zero-sequence sampled values ​​in the same memory area. Bind the two sets of zero-sequence data to the same sampling time timestamp.

[0105] (6) Read the channel compensation status flags attached to each of the three sets of complete phasor sample values. If any one of phases A, B, and C is marked as compensation failure or invalid sample data, then the calculation result of the zero-sequence phasor value of this synthesis is unreliable. The current data is directly stopped from being passed down, and it is determined that S3.2 mutual verification cannot be executed normally.

[0106] S3.2: Calculate the relative amplitude deviation and phase deviation angle between the synthesized zero-sequence sampled value and the corrected measured zero-sequence sampled value. When the relative amplitude deviation is less than the preset first threshold and the phase deviation angle is less than the preset second threshold, the measured zero-sequence sampled value at the current moment is determined to be reliable and the verification is passed.

[0107] Furthermore, the specific steps in S3.2 include:

[0108] (1) Read the synthesized zero-sequence amplitude stored in the temporary memory area of ​​the controller, and at the same time read the actual measured zero-sequence amplitude corresponding to the corrected measured zero-sequence sampling value stored in parallel with the synthesized zero-sequence amplitude. Extract the two sets of amplitudes as the basic data used for the calculation of the relative deviation of amplitude. The synthesized zero-sequence amplitude and the measured zero-sequence amplitude are both calculation results bound to the same GPS microsecond-level sampling timestamp.

[0109] (2) Subtract the measured zero-sequence amplitude from the synthesized zero-sequence amplitude to obtain the difference. Then take the absolute value of the difference without positive or negative, compare the magnitude of the synthesized zero-sequence amplitude and the measured zero-sequence amplitude, select the larger amplitude as the divisor, divide the absolute difference by the larger amplitude, and finally obtain the dimensionless amplitude relative deviation value. The amplitude judgment standard is pre-fixed in the controller program. In this embodiment, the standard is defined as the first threshold. The first threshold is specifically set to 5%, which is the maximum upper limit of the amplitude deviation allowed.

[0110] (3) Read the synthesized zero-sequence phase from the synchronous cache, and at the same time read the measured zero-sequence phase corresponding to the corrected measured zero-sequence sample value. Both the synthesized zero-sequence phase and the measured zero-sequence phase are processed by periodic reduction from 0° to 360°. The original phase difference is obtained by directly subtracting the two sets of phases. When the original phase difference is greater than 180°, the original phase difference is subtracted from 360°. The result is used as the minimum phase deviation angle. The minimum phase deviation angle after processing is fixed between 0° and 180°. The phase judgment standard is pre-fixed in the controller program. In this embodiment, the standard is defined as the second threshold. The second threshold is specifically set to 5°, which is the maximum upper limit of the phase deviation allowed.

[0111] (4) Retrieve the calculated amplitude relative deviation value, compare the amplitude relative deviation value with the first threshold of 5%, and simultaneously retrieve the calculated minimum phase deviation angle, compare the minimum phase deviation angle value with the second threshold of 5°. The two comparison operations are performed simultaneously.

[0112] (5) If the relative amplitude deviation is less than 5% and the minimum phase deviation angle is less than 5°, and both comparison results are met, a mutual verification pass mark is generated, indicating that the current sampling cycle zero-sequence transformer data acquisition data and three-phase current correction results are normal; if the relative amplitude deviation is greater than or equal to 5% or the minimum phase deviation angle is greater than or equal to 5°, and either comparison result does not meet the standard, a mutual verification fail mark is generated, indicating that the synthesized zero-sequence data deviates significantly from the measured zero-sequence data.

[0113] (6) Write the generated mutual verification pass flag or mutual verification fail flag into the internal identifier field of the current frame complete current dataset. This flag serves as the sole criterion for determining whether to output valid current judgment data. If the flag is mutual verification fail, block the transmission of the current frame complete current dataset to S4 and store the entire dataset in the controller's local cache area.

[0114] S3.3: After the verification is passed, the corrected three-phase current sampling value and the corrected measured zero-sequence sampling value are used together to form the effective current judgment data at the current sampling time and output it.

[0115] Furthermore, the mutual verification also includes a fluctuation consistency verification step for the synthesized zero-sequence sampled value: using a continuous preset number of power frequency cycles as a time window, the effective value standard deviation of the synthesized zero-sequence sampled value of each sampling point within the time window is calculated, and the effective value standard deviation of the corrected measured zero-sequence sampled value within the same time window is also calculated. The ratio of the two standard deviations is calculated, and when the ratio is within a preset consistency ratio range, it is determined that the corrected measured zero-sequence sampled value is consistent with the synthesized zero-sequence sampled value in terms of fluctuation trend, further confirming that the measured zero-sequence sampled value at the current moment is reliable, and the verification passes.

[0116] S4: When the preset low load condition is met for the first time, the average value of the DC component of the three-phase current in the effective current judgment data is taken as the baseline bias parameter, and the multi-dimensional state parameters are used together for fusion judgment. The influence weight of each state parameter is dynamically adjusted, and the protection action criterion is iteratively optimized. The protection action criterion is defined as triggering protection when the effective value of any phase in the corrected three-phase current sampling value exceeds 1.2 times the rated value and lasts for more than 40ms.

[0117] The preset low load condition is specifically defined as follows: the circuit breaker is in the closed operation state and the effective value of each phase of the corrected three-phase current sampling value is less than 10% of the rated current and the duration is greater than or equal to 5s; when the low load condition is met for the first time, the arithmetic mean of the DC component sampling points of the three-phase current in the effective current judgment data within one power frequency cycle time window before and after the current time is used as the A-phase baseline offset parameter, B-phase baseline offset parameter and C-phase baseline offset parameter.

[0118] Furthermore, the baseline bias parameter is used to perform real-time zero-drift correction on the corrected three-phase current sample values. Specifically, the A-phase current sample value, B-phase current sample value, and C-phase current sample value acquired at subsequent times are subtracted from the A-phase baseline bias parameter, B-phase baseline bias parameter, and C-phase baseline bias parameter, respectively, to obtain the zero-drift corrected A-phase current sample value, B-phase current sample value, and C-phase current sample value; the zero-drift corrected three-phase current sample values ​​are then used to replace the corrected three-phase current sample values.

[0119] Furthermore, when the preset low load condition is met for the first time, the average value of the DC component of the three-phase current in the effective current determination data is taken as the baseline bias parameter, including:

[0120] (1) The system continuously monitors low load conditions in real time. When all low load conditions are met at the same time, the current sampling time generated by GPS second pulse synchronization with microsecond accuracy is locked as the reference time point. The reference time point is the positioning basis for the data window. The preset low load conditions are fully defined in the previous text as the circuit breaker being in the closed operation state, the effective value of each phase of the three-phase current being less than the rated current, and the duration of this state reaching 5s.

[0121] (2) Using the locked reference time point as the time boundary, retrieve all effective current judgment data within a 20ms interval before the reference time point. 20ms corresponds to a complete cycle of a single power frequency. At the same time, cache the sampling data within a 20ms interval after the reference time point in real time. Merge and integrate the data of the two intervals with a duration of 20ms to form a continuous sampling data window covering two complete power frequency cycles. This data window contains a complete power frequency sampling record before and after the reference time point.

[0122] (3) Read all valid current determination data in the continuous sampling data window that has been integrated in step two. Extract the original sampling values ​​of the DC component of phase A current, the original sampling values ​​of the DC component of phase B current, and the original sampling values ​​of the DC component of phase C current from the data packets corresponding to each frame of valid current determination data. Collect and store the three-phase DC component values ​​corresponding to all sampling points in the window to form a set of DC component values ​​summarized by phase.

[0123] (4) Take out all the DC component sampling values ​​of phase A after collection, calculate the arithmetic mean of all values, and calculate the baseline offset parameter of phase A; use the same calculation method to calculate the arithmetic mean of all DC component sampling values ​​of phase B and phase C after collection, and obtain the baseline offset parameter of phase B and the baseline offset parameter of phase C in turn. The three sets of baseline offset parameters are the core calculation output results of this step.

[0124] (5) Write the calculated A-phase baseline offset parameters, B-phase baseline offset parameters, and C-phase baseline offset parameters into the circuit breaker’s built-in non-volatile memory chip for permanent storage. The three sets of baseline offset parameters serve as the fixed reference for DC zero-point compensation during the long-term operation of the equipment. When reading the DC component values ​​in all current sampling processes of the equipment, the baseline offset parameters stored in this phase need to be deducted accordingly.

[0125] (6) Read the ambient temperature value and the cumulative running time of the equipment from power-on to the reference time point corresponding to the locked reference time point. Simultaneously write the ambient temperature value and the cumulative running time into the non-volatile storage chip, bind and archive it with the three sets of baseline bias parameters, and archive the records for later operation and maintenance to analyze the changes of the zero point value with the ambient temperature and the years of operation of the equipment.

[0126] The joint multi-dimensional state parameters are used for fusion and discrimination, and the influence weight of each state parameter is dynamically adjusted to iteratively optimize the protection action criteria, including:

[0127] S4.1: Construct a comprehensive judgment and calculation link for operation characteristics, with circuit breaker contact temperature, characteristic peak value of operating mechanism opening and closing coil current, ambient temperature and humidity, and cumulative number of circuit breaker operations as input characteristic parameters, and action sensitivity coefficient of protection action criterion as output target;

[0128] Furthermore, the operational feature comprehensive judgment calculation link includes four types of input feature parameter sets: circuit breaker contact temperature, with collection points covering all moving contact temperature measurement points and all stationary contact temperature measurement points of the three phases. The real-time temperature values ​​collected from all temperature measurement points are arithmetically averaged to obtain the average temperature value of the three-phase contacts. Then, the average temperature values ​​of each group of moving and stationary contacts are compared, and the maximum difference between the values ​​is extracted. This maximum difference is used as an auxiliary derived feature corresponding to the contact temperature. The average temperature value of the three-phase contacts and the maximum difference between the three-phase contact temperatures together constitute the first type of input feature; the peak value of the operating mechanism's opening and closing coil current characteristic, during the entire process of each circuit breaker opening or closing operation, all sampled values ​​of the coil circuit current are continuously collected. The maximum value of the current value within a single operation process is extracted as the coil current characteristic peak value. The entire duration of the coil from the start to the end of energization is recorded simultaneously. The coil current characteristic peak value and the coil energization duration together constitute the second type of input feature; ambient temperature and humidity, collected... Two temperature and humidity probes are set up: one at the upper heat dissipation area of ​​the switchgear cabinet and the other at the lower mechanism installation area of ​​the switchgear cabinet. The arithmetic mean of the real-time temperature and humidity values ​​collected by the two probes is taken to obtain the overall ambient temperature and humidity of the cabinet. Multiple sets of continuous overall ambient temperature and humidity data are recorded at fixed sampling intervals. The rate of change of temperature and humidity are calculated by analyzing the recorded values. The overall ambient temperature, overall ambient humidity, rate of change of temperature, and rate of change of humidity together constitute the third type of input feature. The cumulative number of circuit breaker operations is calculated by reading the total number of opening and closing operations of the entire unit stored in the equipment. At the same time, the annual average operation frequency is calculated by counting all operations over a yearly time interval, and the operation frequency for the most recent 30 days is calculated by counting all operations over a fixed 30-day time interval. The cumulative number of circuit breaker operations, the annual average operation frequency, and the operation frequency for the most recent 30 days together constitute the fourth type of input feature.

[0129] Furthermore, the four types of input features are uniformly summarized, and all feature values ​​are sent as pre-input data to the multi-layer linear fusion processing link as pre-processing data. The multi-layer linear fusion processing link consists of three linear calculation stages, with full numerical direct transmission between layers, eliminating data truncation. The first-layer linear calculation stage receives all summarized multi-dimensional input features and assigns a fixed basic weighting coefficient to each group of features. The basic weighting coefficient is the initial weight value obtained by entropy weighting calculation during the equipment manufacturing stage. The first-layer linear calculation stage multiplies each group of feature values ​​with the corresponding basic weighting coefficient and performs a summation operation, outputting the first-layer fusion intermediate value. The first-layer fusion intermediate value is completely transmitted to the second-layer linear calculation stage, which performs a second linear transformation on the first-layer output intermediate value to eliminate the difference in numerical dimensions between different features, outputting the second-layer fusion intermediate value. The second-layer fusion intermediate value is completely transmitted to the third-layer linear calculation stage, which performs the final linear transformation and outputs a normalized single value. This normalized single value is the protection action judgment sensitivity coefficient finally output by the model.

[0130] Furthermore, the original benchmark threshold for protection action judgment is set to 1.2 times the rated current. The magnitude of the action sensitivity coefficient directly corresponds to the relaxation range of the overcurrent trigger threshold. When the action sensitivity coefficient increases, the overcurrent judgment standard is relaxed accordingly, and a load state that slightly exceeds 1.2 times the rated current can be judged as an overload and protection will be activated. When the action sensitivity coefficient decreases, the overcurrent judgment standard is tightened accordingly, and only a load state that continuously and significantly exceeds 1.2 times the rated current will be judged as an overload and protection will be activated. The normalized action sensitivity coefficient output by the multi-layer linear fusion processing link is directly used to adaptively adjust the overcurrent protection judgment threshold that takes effect in real time during the operation of the whole machine.

[0131] S4.2: The feature dimensionality reduction method based on principal component analysis algorithm is used to compress the related features in the multidimensional state parameters into mutually independent principal component features. The principal component analysis algorithm is the prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0132] Furthermore, the specific steps in S4.2 include:

[0133] (1) The controller continuously accumulates historical multidimensional state parameter samples. Each time a complete sampling cycle is completed, a set of feature samples is stored. After the preset number of samples is accumulated, a principal component analysis operation is started.

[0134] (2) Standardize and preprocess all original multidimensional feature samples to eliminate calculation bias caused by different dimensions and numerical ranges of temperature, current and number of operations, and map all features to a unified numerical range.

[0135] (3) Calculate the covariance matrix of all standardized feature samples, and solve for the eigenvalues ​​and corresponding eigenvectors by matrix decomposition. The matrix decomposition is a prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0136] (4) Sort the features in descending order of their values ​​and select all preceding feature vectors whose cumulative contribution accounts for 95% as the principal component transformation basis;

[0137] (5) The original high-dimensional related features are linearly transformed using the transformation basis obtained by screening, and a large number of interrelated original state parameters are compressed into a small number of independent principal component features without linear correlation. After compression, there is no information overlap between the principal component features. After the dimensionality reduction is completed, the principal component features are sent to the input end of the running feature comprehensive judgment operation link in real time to replace the original multidimensional state parameters in weight calculation.

[0138] S4.3: Calculate the initial weight coefficients of each principal component feature in the fusion discrimination based on the entropy weight method;

[0139] Furthermore, the specific steps of S4.3 include:

[0140] (1) Read the historical sample dataset of all independent principal component features output after dimensionality reduction by principal component analysis;

[0141] (2) Calculate the information entropy value of each principal component feature in turn. The size of the information entropy value represents the richness of the effective information for fault discrimination contained in the principal component. The lower the entropy value, the stronger the ability of the feature to distinguish between normal and fault conditions, and the higher the influence weight should be assigned. The higher the entropy value, the weaker the feature's ability to distinguish, and the lower the weight should be assigned. The formula for calculating information entropy is the prior art in this field and is not an inventive solution of this application. It will not be elaborated here.

[0142] (3) Calculate the initial weight allocation ratio based on the information entropy difference of each principal component, normalize the initial weights of all principal components, and unify the sum of all weight values ​​into a fixed total to generate an initial weight coefficient set.

[0143] (4) Bind the normalized initial weight coefficients to the corresponding principal component features and store them in the weight storage area of ​​the running feature comprehensive judgment operation link as the benchmark initial weights before the first iteration optimization after the equipment leaves the factory.

[0144] (5) The initial weight coefficients are backed up synchronously to non-volatile memory and automatically read and loaded each time the whole machine is restarted to ensure that the weight configuration is not lost after power failure.

[0145] S4.4: After each protection action event occurs, record the actual judgment result of the protection action criterion and the actual fault occurrence. When the frequency of false operation increases, backpropagate the weight coefficients of each principal component feature quantity based on the multidimensional state parameter feature value at the time of the false operation event, and gradually reduce the weight coefficients corresponding to the state parameter that contributes the most in the false operation event.

[0146] Furthermore, the specific steps of S4.4 include:

[0147] (1) Two types of abnormal protection events are distinguished. The first type is protection maloperation event: the three-phase current does not reach 1.2 times the rated current and the duration is less than 40ms, but the protection erroneously issues a trip signal. The second type is protection failure to operate event: any one of the three phases exceeds 1.2 times the rated current for more than 40ms for a long time, and the protection does not trigger the trip in time. Both types of events are included in the maloperation frequency statistics. Events that normally and correctly trigger the protection or do not operate are only recorded and do not trigger weight correction.

[0148] (2) When a single false triggering event occurs, retrieve the values ​​of all principal component features corresponding to the time of the event, compare the contribution ratio of each principal component to the sensitivity coefficient under the current weight coefficient, and locate the state parameter corresponding to the principal component with the highest contribution ratio and the greatest impact on this false triggering event.

[0149] (3) Execute the backpropagation weight correction logic, appropriately reduce the principal component weight coefficient corresponding to the high contribution state parameter, and simultaneously slightly increase the weight of the other low interference state parameters. After all weights are completed, normalize again to ensure that the total weight remains unchanged.

[0150] (4) Only a small weight adjustment is made for a single erroneous action to avoid drastic weight fluctuations caused by a single abnormal event. Multiple similar erroneous actions are continuously accumulated and fine-tuned to gradually weaken the influence of state parameters that are prone to misjudgment. The corrected new weight coefficients are temporarily stored in the running memory and wait for the batch update and solidification of the iteration trigger conditions.

[0151] The iterative optimization triggering condition in S4 is:

[0152] B1: When the cumulative number of protection actions reaches the preset batch action number threshold or the cumulative false trip frequency exceeds the preset false trip frequency threshold, a weight coefficient iteration update process is triggered; the false trip frequency is the number of times the protection is falsely triggered even though the criterion conditions are not met, or the protection fails to operate even though the conditions are met.

[0153] Furthermore, the controller is equipped with two independent counters. The first counter is the total number of cumulative protection actions, which increments by 1 for each complete protection action and judgment. The second counter is the false trip frequency counter, which increments by 1 only when a false trip or failure to trip event occurs. The values ​​of both counters are stored in real time and can be read locally and uploaded to the backend. The preset batch action number threshold and false trip frequency threshold can be configured and modified by on-site maintenance personnel through the backend interface. In this embodiment, the factory default batch action threshold is set to 500 protection judgments and the default false trip frequency threshold is set to 10. A complete weight iteration update is triggered when either of the two thresholds is met.

[0154] Furthermore, the total number of protection actions is gradually increased to 500. Regardless of the frequency of false actions, an iterative update is forcibly initiated to periodically optimize the weights and adapt to the changes in state parameter characteristics caused by the long-term operation and aging of the equipment. If the cumulative false action frequency counter value exceeds 10 times but has not reached the total number of actions of 500, an iterative update is also immediately triggered to quickly correct the problem of frequent misjudgments caused by weight reduction. After triggering the iteration, the system synchronously reads all protection event samples, corresponding principal component features, and false action records since the last iteration update. After a single iteration is completed, both sets of counters are synchronously cleared to zero and the cumulative counting starts again to wait for the next iteration to be triggered.

[0155] B2: After each iteration update, the updated weight coefficients are substituted into the operation feature comprehensive judgment calculation link to recalculate the action sensitivity coefficient of the protection action criterion, and the multiple threshold of the rated current in the protection action criterion is adaptively corrected based on the updated action sensitivity coefficient.

[0156] Furthermore, the specific steps of B2 include:

[0157] (1) Complete the calculation work related to the iterative update of the weight coefficients, generate a complete set of new principal component weight coefficients, and load the new principal component weight coefficients into the weight storage area inside the multidimensional running feature comprehensive tuning and judgment process, covering and replacing the old weight configuration content retained before the execution of the iterative operation.

[0158] (2) Retrieve all historical typical working condition samples stored locally on the equipment. The historical typical working conditions include low load working conditions, rated load working conditions, overload working conditions, high temperature operating conditions, and multiple operation working conditions of the mechanism. Send all historical typical working conditions into the multi-dimensional operating feature comprehensive tuning and judgment process with new principal component weight coefficients loaded in sequence, and calculate the new action sensitivity coefficient corresponding to each type of working condition.

[0159] (3) Calculate the arithmetic mean of the action sensitivity coefficients corresponding to all working conditions, and use the calculated average value as the current effective sensitivity benchmark value of the whole machine. Adjust the overload multiple threshold of the protection action criterion based on the current effective sensitivity benchmark value of the whole machine. If the current effective sensitivity benchmark value of the whole machine is too high, adjust the overload multiple threshold slightly upward. If the current effective sensitivity benchmark value of the whole machine is too low, adjust the overload multiple threshold slightly downward. The original benchmark overload multiple of the protection action criterion is set to 1.2 times the rated current.

[0160] (4) Perform upper and lower boundary constraint verification on the adjusted overload multiple threshold. The lower limit of the overload multiple threshold is fixed at 1.1 times the rated current, and the upper limit of the overload multiple threshold is fixed at 1.3 times the rated current. If the adjusted value exceeds the range of 1.1 times to 1.3 times, the threshold will be forcibly corrected to the value closest to the boundary within the range.

[0161] (5) The new principal component weight coefficients and the final overload multiple threshold after boundary constraint verification are synchronously written into the controller's running memory. After the writing is completed, the two sets of parameters take effect in real time and directly participate in the protection judgment calculation work corresponding to each frame current. At the same time, the new principal component weight coefficients and the final overload multiple threshold are synchronously backed up and written into the non-volatile storage medium for permanent retention.

[0162] B3: Repeat the iterative update process until the frequency of false operation is less than or equal to 1 in 100 consecutive protection operations, then freeze the current weight coefficient and action sensitivity coefficient, and stop iterative optimization.

[0163] S5: Based on the optimized protection action criteria, collect current waveform data in real time, connect to the disconnecting switch interlocking signal, and output the current acquisition result after verification and interlocking constraints.

[0164] The specific steps of S5 include:

[0165] S5.1: Under normal operating conditions of the circuit breaker, continuously acquire the secondary current signal of the current transformer of the three-phase bushing of the circuit breaker at a sampling rate of more than 1024 per cycle, generate a real-time current waveform sampling data stream, and obtain the verified real-time valid current sampling sequence after processing by S1-S4.

[0166] Furthermore, when the circuit breaker is operating normally under energized conditions and there is no maintenance tripping command, the controller automatically switches to the high-frequency waveform acquisition mode. In this embodiment, the number of sampling points per cycle is set to 1024 points, and the 50Hz power frequency single cycle is 20ms, which is used to capture the minute current waveform distortion during short circuits and impact overloads, meeting the requirements for fault waveform recording. The analog signals on the secondary side of the three-phase bushing current transformer and the zero-sequence transformer are still processed through the original anti-aliasing filter channel, and the GPS second pulse global synchronization timing is synchronously reused. The four analog signals are synchronously sampled at high frequency, and the analog-to-digital converter continuously outputs high-density digital waveform sampling points at a higher conversion rate, forming a continuous and uninterrupted real-time current waveform sampling data stream. The high-frequency waveform data stream will not skip any processing steps from S1 to S4, and will completely and sequentially perform the original signal preprocessing, layered temperature drift amplitude phase... Position correction, three-phase synthesized zero sequence and measured zero sequence cross-verification, low load baseline offset compensation, multi-source state fusion discrimination optimization protection criterion full-link operation, high-frequency sampling points are grouped and packaged according to cycle. After each group of 1024 waveform data completes the full set of corrections and verifications, a real-time effective current sampling sequence with continuous timing, cross-verification pass mark, and baseline zero-point compensation is generated. The real-time effective current sampling sequence completely retains the original current value, correction compensation parameters, status label, and cross-verification result of each high-frequency sampling point. The sequence is accompanied by a globally unified GPS microsecond-level timestamp to ensure accurate waveform timing and to completely record the instantaneous current change process when a fault occurs. If a current transformer disconnection, filter abnormality, or verification failure is detected during high-frequency acquisition, the corresponding waveform segment is marked with an abnormality mark, and a waveform recording alarm is uploaded synchronously.

[0167] S5.2: The auxiliary contact status signal of the disconnector is acquired in real time through the input interface of the circuit breaker controller; the auxiliary contact status signal includes a first status signal indicating that the disconnector is in the closed position and a second status signal indicating that the disconnector is in the open position.

[0168] Furthermore, the secondary control compartment of the circuit breaker reserves multiple passive input acquisition interfaces, which are electrically connected to the auxiliary contacts of the matching disconnector switch in the same bay via secondary cables. The auxiliary contacts are passive mechanical switching contacts. When the main disconnector switch is closed, the first contact is closed to output the first closed state signal; after the main disconnector switch is fully open, the second contact is switched to close to output the second open state signal; during the intermediate operation transition phase, both contacts are in the open state; the controller's input interface integrates an opto-isolation chip to isolate electromagnetic interference from the high-voltage secondary circuit, continuously and cyclically scanning the level status of each contact, with the scanning period set to milliseconds, to capture the contact opening and closing state transitions in real time; after each scan and reading of the contact level, it is converted into a standardized digital status identifier to distinguish between three working conditions: closed state, open state, and transition without signal. The status identifier is updated every millisecond and synchronously sent to the interlocking logic judgment subroutine, eliminating the problem of status acquisition lag; the acquired auxiliary contact status signals are accompanied by a timestamp, aligned with the current waveform sampling sequence, ensuring that the time base of the interlocking status and current data is consistent.

[0169] S5.3: Perform state discrimination on the acquired auxiliary contact state signal. When the auxiliary contact state signal is determined to be the second state signal, generate a valid flag for the disconnecting switch interlock and set the valid flag to the valid state. When the auxiliary contact state signal is determined to be the first state signal, keep the valid flag for the disconnecting switch interlock in the unset state.

[0170] Furthermore, the interlocking flag setting logic is as follows: continuously receiving millisecond-level updated disconnector switch contact status flags, and completing flag management in three layers of logic. The first layer identifies a stable first state signal, i.e., the disconnector switch is closed in place, and continuously detects that the closing contact has reached the preset anti-jitter duration (20ms in this embodiment) to eliminate instantaneous state misjudgments caused by contact jitter, confirming that the disconnector switch is reliably closed, and the interlocking effective flag remains unset, with the internal register marked as 0. The second layer identifies a stable second state signal, i.e., the disconnector switch is fully open in place, and the open contact has been continuously closed for 20ms of anti-jitter duration, determining that the disconnector switch is in the open-position maintenance condition, and immediately setting the disconnector switch open-position interlocking effective flag to the effective state, with the register marked as 1. The third layer is the operation transition condition where both contacts are simultaneously disconnected. The interlocking flag maintains the value of the previous stable state and does not jump with the instantaneous transition signal, avoiding the frequent flipping of the interlocking flag during the disconnector switch opening and closing operation, which causes the current acquisition sequence to repeatedly be zeroed and restored, resulting in jitter. The interlocking flag value is stored in the global shared register in real time.

[0171] S5.4: When the read disconnector interlocking effective flag is not set, the real-time effective current sampling sequence is directly output to the circuit breaker protection control unit as the effective current acquisition result, so that the circuit breaker protection control unit can perform real-time determination of the optimized protection action criteria.

[0172] Furthermore, when the value of the interlocking valid flag register is not set, it indicates that the primary circuit disconnect switch has been reliably closed and the main circuit has the conditions for normal energized operation. Specifically, this includes: directly reading the processed real-time valid current sampling sequence; adding an interlocking normal operation flag to the head of the sampling sequence to distinguish the interlocking data at the zero position; pushing the complete and unedited data to the back-end circuit breaker protection control unit through the internal parallel data bus of the controller; after receiving the complete valid current sampling sequence, the protection control unit calls the protection action criteria optimized by S4 iteratively to judge in real time whether the three-phase current exceeds 1.2 times the rated current and whether the duration reaches 40ms. If the conditions are met, the trip protection command is immediately output to realize the normal fault protection function; the complete original sampling sequence is synchronously stored locally to the fault recording partition and retained for long-term fault retrospective analysis.

[0173] S5.5: When the interlocking validity flag of the disconnecting switch is set, all sampled values ​​in the real-time valid current sampling sequence at the current moment are forcibly modified to zero, generating a forced zero-current sampling sequence carrying the interlocking action flag. Then, the forced zero-current sampling sequence is output to the circuit breaker protection control unit as a valid current acquisition result to prevent the circuit breaker protection control unit from triggering any protection action based on the forced zero-current sampling sequence.

[0174] Furthermore, when the interlocking flag register is set to an active state, it indicates that the disconnecting switch is open and the primary bay is in a power outage maintenance state. At this time, there is no energized load current in the main circuit. The weak induced noise and leakage current collected by the secondary side of the current transformer will cause the protection to misjudge the overload fault. It is necessary to use the interlocking logic to shield the protection judgment. Specifically, this includes: reading the complete real-time effective current sampling sequence of the current frame, copying the entire structure, timestamp, status label, and cross-check mark of the sequence, and keeping the overall framework of the data packet unchanged; traversing all high-frequency sampling values ​​of each group of three phases A, B, and C and zero sequence in the copied sequence, and uniformly forcibly modifying all current amplitude and phasor values ​​to zero values ​​to completely eliminate the current data caused by the collected noise; in the sequence A dedicated interlocking action marker has been added to the status label field, clearly recording the forced zeroing operation performed by the isolating switch interlock in this frame. The marker includes the precise trigger time of the interlocking zeroing and the confirmation time of the isolating switch opening status. The new sampling sequence with all values ​​set to zero and the interlocking marker is taken as the valid acquisition result for the current period and pushed to the circuit breaker protection control unit. The protection control unit reads the all-zero current sampling sequence. The three-phase current is always zero, which can never meet the 1.2 times rated current overload criterion, completely blocking all overcurrent protection tripping logic to prevent false tripping during maintenance. At the same time, the protection unit recognizes the interlocking action marker and synchronously uploads it to the background monitoring system to prompt the current bay isolating switch opening and protection interlocking closing.

[0175] Furthermore, when the interlocking validity flag of the disconnecting switch is set, a self-verification step for the rationality of the forced zeroing operation is also included: within three consecutive power frequency cycles after the forced zeroing output, the changing trend of the corrected three-phase current sampling values ​​is monitored. If the changing trend shows that the three-phase currents synchronously decay to zero, it is determined that the disconnecting switch state is consistent with the actual current disappearance, the forced zeroing operation is reasonable, and the zeroing output continues to be maintained; if the changing trend shows that the current of one or two phases does not synchronously decay to zero, it is determined that there is an abnormality in the current forced zeroing operation, an interlocking state abnormality alarm signal is issued, and the original real-time valid current sampling sequence without forced zeroing is output as a candidate output result.

[0176] Furthermore, after outputting the current acquisition results verified and interlocked, the method also includes a full-link data traceability and health status self-diagnosis step for the current acquisition results: For each output current acquisition result, key intermediate variables in its entire processing link are associated and recorded. These key intermediate variables include at least the original analog signal amplitude before preprocessing, the fundamental component amplitude after preprocessing, the amplitude temperature drift compensation, the phase temperature drift compensation, the baseline offset parameter value, and the cross-verification deviation value. Based on the recorded key intermediate variables of the entire link, a fault probability inference method based on Bayesian networks is used to calculate the health confidence of the current current acquisition processing link. When the health confidence is lower than a preset health threshold, a link self-check anomaly alarm is output, and a preset number of current acquisition results before the alarm time are marked as data to be reviewed.

[0177] Example 2

[0178] Please see Figure 3 Another embodiment of the present invention provides: a circuit breaker current acquisition and processing system, comprising:

[0179] The signal acquisition module 10 is used to acquire the original analog current signals output by the three-phase bushing current transformer and the zero-sequence current transformer of the circuit breaker, and to simultaneously acquire multi-dimensional status parameters such as circuit breaker contact temperature, operating mechanism coil current, ambient temperature and humidity, and cumulative number of operations; it filters out oversampling half-frequency interference through an anti-aliasing filter, and achieves synchronous sampling of no less than 256 points per cycle by relying on GPS second pulses, obtains digital current signals through analog-to-digital conversion, extracts the fundamental frequency and DC component through fast Fourier transform, encapsulates digital current sampling data packets carrying sampling time, equipment identification status tags, and outputs them outward;

[0180] The layered temperature drift correction module 20 retrieves the amplitude and phase correction coefficient table stored in non-volatile memory, which is calibrated in layers according to temperature range and current amplitude range. It uses the real-time ambient temperature and current amplitude as a two-dimensional search index to match the corresponding correction coefficients. It performs amplitude multiplication coefficient compensation and phase superposition coefficient compensation on the three-phase current and measured zero-sequence current in the data packet, respectively, and then reconstructs the amplitude and phase information, outputting the three-phase current sample value and measured zero-sequence sample value after temperature drift correction.

[0181] The zero-sequence mutual verification module 30 is used to perform phasor summation on the corrected three-phase current at the same time to obtain the synthetic zero-sequence sampled value, calculate the relative amplitude deviation and phase deviation angle between the synthetic zero-sequence sampled value and the measured zero-sequence sampled value, and determine that the verification is passed if both indicators are lower than the corresponding preset thresholds. After the verification is qualified, the corrected three-phase current and the measured zero-sequence current are integrated to form an effective current judgment data output, realizing mutual verification of the two zero-sequence data and eliminating distorted sampled data.

[0182] The criterion optimization module 40 extracts the average value of the DC component of the three-phase current as the baseline bias parameter when the low load condition is met for the first time; it builds a comprehensive judgment calculation link of operation characteristics with various equipment status parameters as input, uses principal component analysis to reduce dimensionality and entropy weight method to assign initial weights to features, and iteratively corrects the weights based on the cumulative protection action and the frequency of false trips, adaptively adjusting the protection action sensitivity coefficient and the protection action criterion of 1.2 times rated current and 40ms delay, and freezes the parameters to stop iteration when there are continuous low false trips.

[0183] The interlocking constraint output module 50 continuously collects real-time current waveforms to generate effective current sampling sequences and reads the interlocking signals of the opening and closing status of the disconnecting switch auxiliary contacts. When the disconnecting switch is closed, it directly outputs the current sampling sequence to the protection unit. When the disconnecting switch is opened, it forces all sampling values ​​to zero and adds an interlocking action mark before outputting them. It relies on the interlocking logic to avoid false triggering of the protection under no-load conditions during the opening.

[0184] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the present invention. All of these variations are within the protection scope of the present invention.

Claims

1. A method for acquiring and processing circuit breaker current, characterized in that, include: S1: Acquire the original analog signals and multi-dimensional status parameters of the circuit breaker during operation, and output a digital current sampling data packet with status label after preprocessing; S2: Based on the pre-stored hierarchical correction coefficients, the amplitude temperature drift compensation and phase temperature drift compensation are performed on the three-phase current and the measured zero-sequence current in the digital current sampling data package, and the corrected three-phase current sampling value and the corrected measured zero-sequence sampling value are output. S3: Calculate the synthesized zero-sequence sampling value based on the corrected three-phase current sampling value, cross-verify the synthesized zero-sequence sampling value with the corrected measured zero-sequence sampling value, and output the effective current judgment data when the verification passes. S4: When the preset low load condition is met for the first time, the average value of the DC component of the three-phase current in the effective current judgment data is taken as the baseline bias parameter, and the multi-dimensional state parameters are used together for fusion judgment. The influence weight of each state parameter is dynamically adjusted, and the protection action criterion is iteratively optimized. The protection action criterion is defined as triggering protection when the effective value of any phase in the corrected three-phase current sampling value exceeds 1.2 times the rated value and lasts for more than 40ms. S5: Based on the optimized protection action criteria, collect current waveform data in real time, connect to the disconnecting switch interlocking signal, and output the current acquisition result after verification and interlocking constraints.

2. The circuit breaker current acquisition and processing method as described in claim 1, characterized in that, The original analog signal includes analog current signals from the secondary side of the current transformer on the three-phase bushing side of the circuit breaker and analog current signals from the secondary side of the zero-sequence current transformer; the multi-dimensional state parameters include at least the circuit breaker contact temperature, the operating mechanism opening and closing coil current, the ambient temperature and humidity, and the cumulative number of circuit breaker operations. The preprocessing specifically includes: filtering the original analog signal using an anti-aliasing filter to remove frequency components higher than 1 / 2 of the sampling frequency; synchronously sampling the filtered analog signal based on the GPS second pulse signal, with the sampling frequency set to no less than 256 points per cycle; converting the synchronously sampled signal into a digital current sampling signal using an analog-to-digital converter; performing a fast Fourier transform on the digital current sampling signal to extract the fundamental and DC components; encapsulating the extracted fundamental and DC components into a digital current sampling data packet based on a preset sampling data packet format, and attaching a status label containing the sampling time and circuit breaker identification to the digital current sampling data packet.

3. The circuit breaker current acquisition and processing method as described in claim 2, characterized in that, The hierarchical correction coefficients are pre-stored in non-volatile memory. The hierarchical correction coefficients are calibrated hierarchically according to the ambient temperature range and the current amplitude range. Specifically, during the circuit breaker factory calibration, standard current signals are injected through a signal source for multiple preset ambient temperature ranges and multiple preset current amplitude ranges. The amplitude deviation and phase deviation between the sampled measurement value and the standard injected value are calculated. Based on the amplitude deviation and phase deviation, an amplitude correction coefficient table and a phase correction coefficient table corresponding one-to-one with each temperature range and each current amplitude range are generated. When applying the correction coefficients, the ambient temperature value is used as the first search index to determine the calibration temperature range to which the current temperature belongs, and the current amplitude value is used as the second search index to determine the preset current amplitude range to which the current current belongs. The applicable amplitude correction coefficients and phase correction coefficients at the current moment are obtained from the amplitude correction coefficient table and the phase correction coefficient table through a two-dimensional lookup table.

4. The circuit breaker current acquisition and processing method as described in claim 3, characterized in that, The specific steps of S2 include: Based on the ambient temperature value in the multidimensional state parameters at the current moment, determine the calibration temperature range to which the current temperature belongs, call the amplitude correction coefficient under the calibration temperature range, and multiply the amplitudes of the three-phase current and the measured zero-sequence current in the digital current sampling data packet by the corresponding amplitude correction coefficients respectively to obtain the amplitude-compensated three-phase current amplitude and the measured zero-sequence current amplitude. Based on the ambient temperature value in the multidimensional state parameters at the current moment, determine the calibration temperature range to which the current temperature belongs, call the phase correction coefficient under the calibration temperature range, and superimpose the corresponding phase correction coefficients on the phases of the three-phase current and the measured zero-sequence current in the digital current sampling data packet to obtain the phase-compensated three-phase current phase and the measured zero-sequence current phase. The amplitude of the three-phase current after amplitude compensation is combined with the phase of the three-phase current after phase compensation to reconstruct the corrected three-phase current sample value. The measured zero-sequence current amplitude after amplitude compensation is combined with the measured zero-sequence current phase after phase compensation to reconstruct the corrected measured zero-sequence sampling value.

5. The circuit breaker current acquisition and processing method as described in claim 4, characterized in that, The specific steps of S3 include: The A-phase current sample value, B-phase current sample value, and C-phase current sample value of the corrected three-phase current sample values ​​at the same sampling time are summed by phasor operation to obtain the composite zero-sequence phasor value at that sampling time. Calculate the relative amplitude deviation and phase deviation angle between the synthesized zero-sequence sampled value and the corrected measured zero-sequence sampled value. When the relative amplitude deviation is less than a preset first threshold and the phase deviation angle is less than a preset second threshold, the measured zero-sequence sampled value at the current moment is determined to be reliable and the verification is passed. After the verification is passed, the corrected three-phase current sampling value and the corrected measured zero-sequence sampling value are used together to form the effective current determination data at the current sampling time and output it.

6. The circuit breaker current acquisition and processing method as described in claim 5, characterized in that, The preset low load condition is specifically defined as follows: the circuit breaker is in the closed operation state and the effective value of each phase of the corrected three-phase current sampling value is less than 10% of the rated current and the duration is greater than or equal to 5s; when the low load condition is met for the first time, the arithmetic mean of the DC component sampling points of the three-phase current in the effective current judgment data within one power frequency cycle time window before and after the current time is used as the A-phase baseline offset parameter, B-phase baseline offset parameter and C-phase baseline offset parameter.

7. The circuit breaker current acquisition and processing method as described in claim 6, characterized in that, The joint multi-dimensional state parameters are used for fusion and discrimination, and the influence weight of each state parameter is dynamically adjusted to iteratively optimize the protection action criteria, including: A comprehensive operational characteristic judgment calculation link is constructed, with circuit breaker contact temperature, characteristic peak value of operating mechanism opening and closing coil current, ambient temperature and humidity, and cumulative number of circuit breaker operations as input characteristic parameters, and action sensitivity coefficient of protection action criterion as output target. A feature dimensionality reduction method based on principal component analysis algorithm is used to compress the relevant features in multidimensional state parameters into mutually independent principal component features. The initial weight coefficients of each principal component feature in the fusion discrimination are calculated based on the entropy weight method. After each protection action event occurs, the actual judgment result of the protection action criterion and the actual fault occurrence are recorded. When the frequency of false operation increases, the weight coefficients of each principal component feature quantity are backpropagated and corrected based on the multidimensional state parameter feature values ​​at the time of the false operation event.

8. The circuit breaker current acquisition and processing method as described in claim 7, characterized in that, The iterative optimization triggering condition in S4 is: when the cumulative number of protection actions reaches the preset batch action number threshold or the cumulative false operation frequency exceeds the preset false operation frequency threshold, a weight coefficient iterative update process is triggered. After each iteration, the updated weight coefficients are substituted into the operation feature comprehensive judgment calculation link to recalculate the action sensitivity coefficient of the protection action criterion, and the multiple threshold of the rated current in the protection action criterion is adaptively corrected based on the updated action sensitivity coefficient. Repeat the iterative update process until the frequency of false trips within 100 consecutive protection operations is less than or equal to 1. Then freeze the current weight coefficient and action sensitivity coefficient and stop the iterative optimization. The frequency of false trips is the number of times the protection is falsely triggered when the criterion conditions are not met or when the conditions are met but the protection fails to operate.

9. The circuit breaker current acquisition and processing method as described in claim 8, characterized in that, The specific steps of S5 include: Under normal operating conditions of the circuit breaker, the current signal of the secondary side of the current transformer of the three-phase bushing of the circuit breaker is collected, a real-time current waveform sampling data stream is generated, and the verified real-time effective current sampling sequence is obtained. The auxiliary contact status signal of the disconnecting switch is acquired in real time through the input interface of the circuit breaker controller; the auxiliary contact status signal includes a first status signal indicating that the disconnecting switch is in the closed position and a second status signal indicating that the disconnecting switch is in the open position. The acquired auxiliary contact status signal is judged. When the auxiliary contact status signal is judged to be the second status signal, a valid flag for the disconnecting switch interlock is generated and the valid flag is set to the valid state. When the auxiliary contact status signal is judged to be the first status signal, the valid flag for the disconnecting switch interlock is kept in the unset state. When the read disconnector interlocking effective flag is not set, the real-time effective current sampling sequence is directly output to the circuit breaker protection control unit as the effective current acquisition result. When the read disconnector interlocking effective flag is in the set state, all sampled values ​​in the real-time effective current sampling sequence at the current moment are forcibly modified to zero, generating a forced zero-set current sampling sequence carrying the interlocking action flag, and then the forced zero-set current sampling sequence is output to the circuit breaker protection control unit as an effective current acquisition result.

10. A circuit breaker current acquisition and processing system, used to implement the circuit breaker current acquisition and processing method according to any one of claims 1-9, characterized in that, include: The signal acquisition module is used to acquire the original analog current signal and multi-dimensional state parameters of the circuit breaker; The layered temperature drift correction module uses real-time ambient temperature and current amplitude as a two-dimensional search index to match the corresponding correction coefficients. It performs amplitude multiplication coefficient compensation and phase superposition coefficient compensation on the three-phase current and the measured zero-sequence current respectively, and then reconstructs the amplitude and phase information, outputting the three-phase current sampled value and the measured zero-sequence sampled value with temperature drift correction completed. The zero-sequence mutual verification module is used to sum the phasors of the corrected three-phase currents at the same time to obtain the composite zero-sequence sampled value, calculate the relative amplitude deviation and phase deviation angle between the composite zero-sequence sampled value and the measured zero-sequence sampled value, and generate effective current judgment data output after the verification is qualified. The criterion optimization module is used to extract the average value of the DC component of the three-phase current as the baseline bias parameter, build a comprehensive judgment calculation link for operation characteristics with various equipment status parameters as input, and iteratively correct the protection action sensitivity coefficient and protection action criterion based on the cumulative protection action and the frequency of false action triggering weights. The interlocking constraint output module collects real-time current waveforms to generate effective current sampling sequences, reads the interlocking signals of the opening and closing status of the disconnecting switch auxiliary contacts, directly outputs the current sampling sequence to the protection unit when the disconnecting switch is closed, and forces all sampled values ​​to zero and adds interlocking action markers before outputting them when the disconnecting switch is open.

Images