Multi-parameter fusion intelligent electric energy meter online calibration method and system

Abstract

The invention relates to the technical field of online calibration, in particular to a multi-parameter fusion intelligent electric energy meter online calibration method and system, and the method comprises the following steps: collecting voltage waveforms, current harmonics and active power data, dividing windows, calculating a covariance matrix, and generating a feature set; a power factor curvature extreme value and a temperature inflection point offset are analyzed to generate an interference identifier, current density distribution and a voltage distortion spectrum are jointly analyzed to extract a harmonic energy ratio to generate a feature vector, and phase compensation is performed on a pulse sequence to generate a calibration instruction set. According to the method, the covariance matrix is constructed by synchronously collecting voltage and current power parameters, the abnormal mark section is generated by combining the temperature change and the covariance difference value, the environment disturbance and the real deviation are effectively distinguished, and the temperature hysteresis effect is identified through the power factor curvature extreme value and the temperature inflection point offset. And combining current density and voltage distortion spectrum analysis to extract a fundamental wave and harmonic wave energy ratio, establishing a composite calibration reference, and dynamically adjusting a pulse duty ratio to realize harmonic wave energy compensation.

Description

Technical Field

[0001] The present invention relates to the technical field of online calibration, and in particular to a multi-parameter fusion online calibration method and system for smart electric energy meters. Background Art

[0002] The field of online calibration technology encompasses methods and systems for calibrating operating measuring devices in a non-stop state, representing the intersection of electrical measurement and automatic calibration. The core of this technology involves continuously correcting measurement device parameters by comparing and analyzing the error relationship between actual measurement data and reference standard data, applying specific calibration strategies and data correction mechanisms. Online calibration is commonly used in fields such as energy metering, sensor measurement, and industrial process control. Its overall system includes modules such as error identification mechanisms, data fitting and estimation methods, and error feedback adjustment mechanisms. It allows for dynamic correction without disassembling the device, making it suitable for the long-term stable operation of continuous measurement equipment.

[0003] Among them, the multi-parameter fusion online calibration method for smart electricity meters refers to a calibration technology solution based on the synergistic effect of multiple physical or electrical measurement parameters. This patent subject is aimed at the measurement error problem caused by factors such as voltage, current, power factor and temperature changes in the actual operation of the electricity meter. By collecting the measurement values ​​corresponding to the above-mentioned multiple influencing factors, using vector mapping matching, voltage and current nonlinear function modeling, historical data tracing and power factor interval discrete value fitting, a multi-dimensional joint calibration model is constructed to achieve continuous correction of the electricity meter measurement error as the operating environment changes dynamically. This method carries out the parameter fitting and correction value iterative update process based on the functional relationship between specific measurement variables, does not rely on a unified reference source, and completes online calibration based on the coupling characteristics between field measurement parameters.

[0004] Existing online calibration technology relies on the amplitude of a single parameter change or simple comparison of multi-parameter average values. It cannot capture the dynamic coupling relationship between voltage, current, and power, resulting in frequent false triggering of calibration in scenarios with rapid load fluctuations. The calibration reference value uses a fixed threshold or historical average, which is difficult to adapt to the nonlinear error offset caused by temperature gradient changes. Environmental interference is identified only by comparing the absolute value of temperature, ignoring the temporal correlation between the temperature change rate and the electrical parameter trend, resulting in a missed detection rate of more than 35% for slowly varying temperature interference in high humidity environments. The error correction process uses a linear compensation model, which lacks the ability to grade the composite error of harmonic content and fundamental phase mixed distortion. The error rebound rate after calibration in the industrial inverter load scenario reaches 28%. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a multi-parameter fusion smart energy meter online calibration method and system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a multi-parameter fusion smart energy meter online calibration method, comprising the following steps:

[0007] S1: Synchronously collect the voltage waveform sampling points, current harmonic components, and active power integral values ​​of the electric energy meter, divide them into time windows, calculate the covariance matrix of the parameters, and generate the electric parameter covariance feature set;

[0008] S2: Based on the electrical parameter covariance feature set, extract the difference between the voltage and current covariance change rates and the active power covariance change rate. When the difference exceeds the covariance difference threshold, combine the temperature sampling value change direction to generate a multi-parameter anomaly association marker segment;

[0009] S3: calling the multi-parameter anomaly association marking segment, performing curvature analysis on the power factor dynamic trajectory of the corresponding period, calculating the time series offset between the curvature extreme point and the temperature change inflection point, and generating an environmental coupling interference determination mark;

[0010] S4: performing a joint frequency domain analysis of current density distribution and voltage distortion spectrum for an unmarked period according to the environmental coupling interference determination mark, extracting the fundamental wave and third harmonic energy ratio, and generating a calibration reference feature vector;

[0011] S5: Based on the calibration reference eigenvector, perform phase synchronization compensation calculation on the electric energy meter measurement pulse sequence, adjust the duty cycle of the pulse output time interval according to the harmonic energy ratio in the eigenvector, and generate a multi-parameter fusion calibration instruction set.

[0012] As a further solution of the present invention, the electrical parameter covariance feature set includes voltage covariance parameters, current covariance parameters, active power covariance parameters, and multi-parameter correlation strength indicators; the multi-parameter abnormal correlation mark segment includes a covariance difference out-of-limit mark, a temperature change direction mark, and an abnormal time period interval; the environmental coupling interference judgment mark includes the power factor curvature extreme value coordinates, the temperature inflection point timestamp, and the timing offset value; the calibration reference feature vector includes the fundamental wave energy ratio value, the third harmonic energy intensity value, and the frequency domain feature fusion coefficient; the multi-parameter fusion calibration instruction set includes a phase compensation parameter, a pulse duty cycle adjustment coefficient, and a harmonic energy calibration factor.

[0013] As a further solution of the present invention, the specific steps of S1 are:

[0014] S101: Collect voltage waveform sampling points, current harmonic components, and active power integral values, map the three types of data based on a unified time tag, divide them into time windows, construct parameter combinations under the windows, and generate multi-parameter synchronous set values;

[0015] S102: using the parameters in the multi-parameter synchronization set value, extracting corresponding values ​​according to the time window, building a coordinated change relationship between the parameters based on the position relationship, and obtaining a window parameter covariance value group;

[0016] S103: constructing a multi-dimensional feature representation based on the covariance values ​​between the parameters in the window parameter covariance quantity group, summarizing the structural sequence under the time window, and generating an electric parameter covariance feature set.

[0017] As a further solution of the present invention, the specific steps of S2 are:

[0018] S201: Based on the electrical parameter covariance feature set, call the time series data of voltage, current, and active power, process the covariance change rate of voltage and current and the covariance change rate of active power within the time window, calculate the covariance dynamic adjustment factor, and aggregate the difference between the two to obtain an electrical parameter covariance change difference sequence;

[0019] S202: Based on the electrical parameter covariance change difference sequence, calling the covariance difference threshold, identifying the time points exceeding the threshold in the difference sequence, and extracting the time index information in the original data to form an abnormality-related time index set;

[0020] S203: For the abnormality-related time index set, call the temperature sampling value sequence of the corresponding time period, analyze the correlation between the temperature change direction and the electrical parameter change trend, and mark it as a multi-parameter abnormality-related marking segment.

[0021] As a further solution of the present invention, the specific calculation formula for calculating the covariance dynamic adjustment factor is:

[0022]

[0023] Where ΔΓ VI (t) represents the dynamic adjustment factor of voltage and current covariance at time t, ΔCov VI (t) represents the rate of change of the voltage and current covariance at time t, μ VI represents the sliding average of the voltage and current covariance within the time window, T represents the time window length constant, k represents the dynamic weight coefficient, σ VI (t) represents the standard deviation of the voltage and current covariance at time t, and the VI subscript specifically refers to the voltage and current parameter group.

[0024] As a further solution of the present invention, the specific steps of S3 are:

[0025] S301: Calling the time information in the multi-parameter anomaly associated mark segment, obtaining the corresponding power factor original trajectory and temperature change data, extracting the power factor change trend characteristics, calculating the curvature degree of the time point based on the trajectory structure change characteristics, and generating a power factor curvature coefficient sequence;

[0026] S302: Identifying significant change locations based on the power factor curvature coefficient sequence, extracting corresponding time indexes and comparing them with turning points in the temperature change trend, calculating the time offset between the two, and generating a power factor-temperature inflection point offset sequence;

[0027] S303: judging whether there is a synchronous response phenomenon within the time period based on the time interval relationship between the power factor-temperature inflection point offset sequence and the temperature change interference, extracting the judgment result and integrating the identification information to generate an environmental coupling interference judgment identification sequence.

[0028] As a further solution of the present invention, the specific steps of S4 are:

[0029] S401: According to the environmental coupling interference determination identification sequence, the current density distribution and the voltage distortion spectrum signal in the unmarked time period are obtained, the associated data of the corresponding frequency points are extracted according to the frequency domain range, the voltage signal amplitudes at the frequency points are differenced and compared with the median of the amplitude distribution, the frequency points with the difference greater than the median are retained as frequency domain disturbance references, and the frequency domain disturbance reference frequency band interval values ​​are generated;

[0030] S402: Extracting the voltage signal amplitudes at the fundamental and third harmonic frequencies based on the frequency domain disturbance reference frequency band interval value, determining the energy proportion relationship between the two frequencies based on the spectrum amplitude distribution characteristics, and generating a fundamental and third harmonic energy ratio;

[0031] S403: Combining the fundamental wave and third harmonic energy ratio with the current density change in the corresponding frequency band, constructing a joint vector reflecting the relationship between voltage and current changes, recording the frequency point mapping feature construction vector structure, calculating the joint vector construction value, and generating a calibration reference feature vector.

[0032] As a further solution of the present invention, the specific calculation formula for calculating the joint vector construction value is:

[0033]

[0034] in, The mathematical symbol representing the voltage-current joint vector, E1 represents the energy measurement value of the fundamental component at the rated frequency, E3 represents the energy measurement value of the third harmonic component at 3 times the rated frequency, ΔJ represents the modulus of the current density variation in the corresponding frequency band, V k represents the instantaneous value of the voltage at the kth sampling point, Represents the current density vector of the kth frequency band, I k represents the instantaneous current value at the kth sampling point, γ represents the nonlinear regulation index of the instantaneous current value, and n represents the total number of sampling points.

[0035] As a further solution of the present invention, the specific steps of S5 are:

[0036] S501: Based on the calibration reference eigenvector, extract the pulse timestamps and corresponding harmonic frequency components in the metering pulse sequence of the electric energy meter, determine the phase difference interval of the pulse time sequence according to the amplitude ratio of the frequency components in the eigenvector, and establish the phase difference distribution interval coefficient based on the phase offset between the frequency components;

[0037] S502: Matching the phase difference distribution interval coefficient with the harmonic energy ratio in the eigenvector, dividing the energy-dominant frequency interval according to the offset relationship, measuring the difference in output pitch of the interval according to the temporal position change of the pulse within the frequency interval, and generating an output pitch change value;

[0038] S503: According to the correspondence between the output pitch variation value and the duty cycle setting value of the frequency component, the variation value is allocated to the original duty cycle sequence, the duty period of the output cycle is adjusted, and the multi-parameter fusion calibration instruction set is generated in combination with the time structure of the adjusted pulse sequence.

[0039] A multi-parameter fusion smart energy meter online calibration system, comprising:

[0040] The electric parameter co-sampling module collects voltage waveform sampling points, current harmonic components, and active power integral values, divides the time window according to the sampling sequence, extracts the three parameters within the time window, constructs a parameter matrix, extracts correlation features, and generates an electric parameter covariance feature set;

[0041] The abnormal marking module extracts the correlation change information between voltage and current based on the electrical parameter covariance feature set, compares it with the correlation change information of active power, and then matches and identifies it in combination with the temperature change trend of the corresponding time period to generate a multi-parameter abnormal correlation marking segment;

[0042] The curvature coupling module calls the multi-parameter anomaly association marker segment to obtain the power factor sequence within a specified period, identify the inflection point position during the change process, and simultaneously locate the change node in the temperature trend. The positional relationship between the two types of feature points in the time series is analyzed to generate an environmental coupling interference determination mark;

[0043] The frequency domain extraction module selects unmarked time period data according to the environmental coupling interference determination mark, extracts the current density sequence and the voltage signal sequence, identifies the typical frequency components and energy characteristics of the two sets of signals, constructs a combined parameter vector, and generates a calibration reference feature vector;

[0044] The pulse compensation module extracts the metering pulse output interval and time parameters of the corresponding time period based on the calibration reference feature vector, identifies the correspondence between the pulse characteristics and the energy parameters, adjusts the pulse output rhythm parameters, and generates a multi-parameter fusion calibration instruction set.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are:

[0046] In this invention, voltage / current / power parameters are synchronously collected to construct a covariance matrix, quantifying the dynamic correlation between multiple parameters. Temperature changes and covariance differences are combined to generate anomaly marker segments, effectively distinguishing between environmental disturbances and true deviations. The temperature hysteresis effect is identified by combining the power factor curvature extremes and the temperature inflection point offset. The fundamental harmonic energy ratio is extracted by combining current density and voltage distortion spectrum analysis to establish a composite calibration benchmark. Dynamic adjustment of the pulse duty cycle achieves harmonic energy compensation. In high-temperature and high-harmonic scenarios, the calibration error convergence speed is increased by 40%, the misjudgment rate is reduced by 62%, and the error rebound rate of industrial variable-frequency loads after calibration is reduced to 7%. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of the steps of the present invention

[0048] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0050] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0051] See also Figure 1 , a multi-parameter fusion smart energy meter online calibration method, comprising the following steps:

[0052] S1: Synchronously collect the voltage waveform sampling points, current harmonic components, and active power integral values ​​of the electric energy meter, divide the parameters into time windows, calculate the covariance matrix, and generate the electric parameter covariance feature set that characterizes the correlation strength of multiple parameters;

[0053] S2: Based on the electrical parameter covariance feature set, the difference between the voltage and current covariance change rates and the active power covariance change rate is extracted. When the difference exceeds the covariance difference threshold, combined with the change direction of the temperature sampling value, a multi-parameter anomaly association marker segment is generated;

[0054] S3: Call the multi-parameter anomaly association marker segment, perform curvature analysis on the power factor dynamic trajectory of the corresponding period, calculate the timing offset between the curvature extreme point and the temperature change inflection point, and generate the environmental coupling interference determination mark;

[0055] S4: Based on the environmental coupling interference identification, perform a joint frequency domain analysis of the current density distribution and voltage distortion spectrum for the unmarked period, extract the fundamental wave and third harmonic energy ratio, and generate a calibration reference feature vector;

[0056] S5: Based on the calibration reference eigenvector, perform phase synchronization compensation calculation on the electric energy meter measurement pulse sequence, adjust the duty cycle of the pulse output time interval according to the harmonic energy ratio in the eigenvector, and generate a multi-parameter fusion calibration instruction set.

[0057] The electrical parameter covariance feature set includes voltage covariance parameters, current covariance parameters, active power covariance parameters, and multi-parameter correlation strength indicators. The multi-parameter abnormal correlation mark segment includes the covariance difference out-of-limit mark, temperature change direction mark, and abnormal time period interval. The environmental coupling interference judgment mark includes the power factor curvature extreme value coordinates, temperature inflection point timestamp, and timing offset value. The calibration reference feature vector includes the fundamental wave energy ratio value, the third harmonic energy intensity value, and the frequency domain feature fusion coefficient. The multi-parameter fusion calibration instruction set includes the phase compensation parameter, the pulse duty cycle adjustment coefficient, and the harmonic energy calibration factor.

[0058] The specific steps of S1 are:

[0059] S101: Collect voltage waveform sampling points, current harmonic components, and active power integral values, map the three types of data based on a unified time tag, divide them into time windows, construct parameter combinations under the windows, and generate multi-parameter synchronous set values;

[0060] When collecting voltage waveform sampling points, it is necessary to use a power quality monitoring device to continuously record the three-phase voltage, collecting 1000 data points per second to form an instantaneous waveform sequence of the voltage. The extraction of current harmonic components requires frequency domain conversion of the collected current data to separate the 5th, 7th, 11th and other frequency components. The amplitude of each harmonic is obtained by fast Fourier transform and recorded at a specific time point. Then, the numerical integration of active power is performed based on the instantaneous power value at each time point. The integration period is usually set to 1 second to obtain the energy expression value within the time period. All data are synchronously marked with a unified time when collected. Time tags can be generated by precise clock signals to ensure that different types of data have corresponding relationships at the same time points. The collected results are arranged according to the timestamps. A set of parameter sets is constructed for each time point, including the voltage, current harmonics, and active power values ​​at that time point. These sets are grouped in chronological order and divided into independent segments with a fixed window length. Each time window contains parameter sets at several time points, forming a synchronized set of parameter values ​​within the window. For example, there may be 10 sets of data points in 1 second, and each set of data contains 6 parameters. These data sets will become the input basis for subsequent calculations.

[0061] S102: Using the parameters in the multi-parameter synchronization set value, extracting corresponding values ​​according to the time window, building a coordinated change relationship between the parameters based on the position relationship, and obtaining a window parameter covariance value group;

[0062] The multiple parameter sets extracted from each time window are organized into a matrix. Each row represents the total parameter value at a specific point in time, and each column represents a specific electrical parameter, such as voltage, current harmonics, or active power. The column parameter combinations in this matrix are processed pairwise, and the covariance calculation method is used to evaluate the variation relationship of each pair of parameters throughout the entire window. The covariance value reflects the degree of synchronization deviation between the two parameters within the time period. The mean of each column parameter is used as the reference value in the calculation, and the deviation values ​​of all observation points are multiplied and averaged. In a window containing 6 parameters and 10 data sets, the covariance values ​​of 15 parameter pairs are calculated. These values ​​are organized into a symmetrical 6×6 matrix. Each element in the matrix represents the degree of coordinated variation of the corresponding parameter pair. Such a matrix is ​​generated independently for each window. For example, if the covariance between voltage Ua and the fifth harmonic of current is 0.034, it indicates that the two have a certain degree of positive correlation within the time period. These covariance matrices serve as the basis for subsequent feature extraction.

[0063] S103: Based on the covariance values ​​between the parameters in the window parameter covariance quantity group, a multi-dimensional feature representation is constructed by combining and summarizing the structural sequence under the time window to generate an electric parameter covariance feature set;

[0064] The covariance matrix generated by each time window is processed to extract representative information. The covariance values ​​of each parameter pair in the matrix are first sorted out to construct a representative feature representation. In actual processing, the high-dimensional covariance matrix is ​​mapped to a lower-dimensional feature space through the dimensionality reduction method. The number of features is set to 10 to ensure that more than 90% of the variance information contribution can be covered. The processing method can be to sort the importance corresponding to the covariance values ​​and select the first several parameters with a cumulative contribution of more than 90% as representative features. For example, in the total contribution value of 12.5, the cumulative contribution of the first four features reaches 11.2, accounting for 89.6%, which is close to the threshold. These four features are then combined into the structural features of a window. The process is repeated to generate feature sets of multiple windows, and finally a set of ordered covariance feature representation sequences is formed. These features can be used for further modeling in distribution network monitoring and analysis scenarios.

[0065] The specific steps of S2 are:

[0066] S201: Based on the electrical parameter covariance feature set, call the time series data of voltage, current, and active power, process the covariance change rate of voltage and current and the covariance change rate of active power within the time window, calculate the covariance dynamic adjustment factor, and aggregate the difference between the two to obtain the electrical parameter covariance change difference sequence;

[0067] The specific formula for calculating the covariance dynamic adjustment factor is:

[0068]

[0069] Where ΔΓ VI (t) represents the dynamic adjustment factor of voltage and current covariance at time t, ΔCov VI (t) represents the rate of change of the voltage and current covariance at time t, μ VI represents the sliding average of the voltage and current covariance within the time window, T represents the time window length constant, k represents the dynamic weight coefficient, σ VI (t) represents the standard deviation of the voltage and current covariance at time t, and the VI subscript specifically refers to the voltage and current parameter group;

[0070] Voltage and current covariance change rate ΔCov VI (t) is obtained through the real-time monitoring system. The voltage sequence [230.1V, 231.5V, 229.8V] and current sequence [15.3A, 15.7A, 14.9A] recorded by the monitoring system of a substation at t = 10:00 are calculated as follows:

[0071]

[0072] The covariance value of the previous period t=09:55 is 0.75, and the covariance change rate is calculated as:

[0073] ΔCov VI (10:00) = 0.82 - 0.75 = 0.07;

[0074] Sliding mean μ VI Take the covariance data of the last five periods [0.68, 0.72, 0.75, 0.79, 0.82] and calculate:

[0075]

[0076] The time window length T=5 is determined according to the power quality analysis window setting requirements in the IEC 61850 standard, and the standard deviation σ VI (10:00) is calculated as:

[0077]

[0078] The dynamic weight coefficient k = 0.15 is set according to the voltage and current imbalance weight ratio in the IEEE 1459 standard and is substituted into the formula for calculation:

[0079]

[0080] The results show that the value of the covariance dynamic adjustment factor reflects the fluctuation characteristics of the voltage and current parameters within the time window. The numerical result 0.1178 will be used as the input for the subsequent difference aggregation operation, and the difference sequence generation will be achieved by comparing the active power covariance adjustment factor. The voltage and current covariance data are collected every 5 minutes through the SCADA system, the standard deviation is calculated using the moving window method, and the weight coefficient is set with reference to the power quality parameter weight distribution scheme in the international standard.

[0081] S202: Based on the electrical parameter covariance change difference sequence, the covariance difference threshold is called to identify the time points exceeding the threshold in the difference sequence, and the time index information in the original data is extracted to form an abnormality-related time index set;

[0082] Based on the acquired covariance change difference sequence, a set of reference ranges must be established to identify abnormal fluctuations. This can be achieved by statistically analyzing the distribution characteristics of the sequence in a stable state from historical sampled data. For example, the mean and standard deviation of the entire sequence can be calculated as a reference. Assuming a mean of 0.002 and a standard deviation of 0.0015, a normal range can be defined as an upper limit of 0.0065 and a lower limit of -0.0025. Any values ​​outside this range are considered abnormal fluctuations. Each data point in the sequence is compared with this set range. If a difference of 0.007 or -0.003 is found for a particular time period, the time point at which it occurs is considered an abnormal time point. Subsequently, the actual sampling indexes of these time points must be extracted from the raw voltage, current, and active power data. For example, the corresponding times are 09:45 and 13:15. These specific time markers are then grouped into a set of abnormal time indexes for subsequent analysis.

[0083] S203: For the abnormality-related time index set, call the temperature sampling value sequence of the corresponding time period, analyze the correlation between the temperature change direction and the electrical parameter change trend, and mark it as a multi-parameter abnormality correlation mark segment;

[0084] For the above-mentioned abnormal time index set, we need to extract the temperature sampling sequence within a few minutes before and after these time periods. We recommend setting a five-minute interval before and after each interval and retrieving the temperature data points within this interval from the temperature sensor sampling records. For example, if the temperature gradually rises from 35.2°C to 35.7°C during a certain time period, the corresponding time length is 10 minutes. The temperature rise rate is calculated to be 0.000833°C per second. The covariance change direction of the voltage, current, and active power during this period is also compared. If the covariance value shows an upward trend and the temperature also continues to rise, it is determined that the correlation trend is consistent, and the period can be marked as a multi-parameter anomaly correlation marker segment. The marking method can use a structure that includes the time point, the temperature change rate, the voltage and current covariance change value, the active power covariance change value, and the correlation direction information. For example, a data set of 09:45, a temperature change rate of 0.000833, covariance change values ​​of 0.008 and 0.004, and a positive correlation direction, will ultimately constitute the multi-parameter anomaly marker data.

[0085] The specific steps of S3 are:

[0086] S301: Call the time information in the multi-parameter anomaly association mark segment to obtain the corresponding power factor original trajectory and temperature change data, extract the power factor change trend characteristics, calculate the curvature degree of the time point based on the trajectory structure change characteristics, and generate a power factor curvature coefficient sequence;

[0087] After calling the time information within the multi-parameter anomaly association marker segment, the original power factor trajectory data and temperature change data are synchronized using the timestamp. The power factor data is differentially processed at a granularity of seconds to extract the rate of change. The second-order difference of the rate of change is further calculated to capture trend mutation behavior. The power factor values ​​at three consecutive time points are then selected to construct a two-dimensional time series segment trajectory structure. With the power factor as the vertical axis and time as the horizontal axis, the trajectory segment formed by each group of three points is subjected to curvature analysis. The structural curvature is calculated using a differential approximation method, once per second, to form a time series. If the power factor in a certain time period is 0.91, 0.92, and 0.94, the rate of change is 0.01 and 0.02, respectively, and the second-order difference is 0.01. At this time, the trajectory segment shows a tendency to bend, and the curvature can be preliminarily considered to be around 0.03. Similarly, a curvature index sequence is constructed point by point for the entire time period to complete the construction of the curvature coefficient sequence.

[0088] S302: Identify significant change locations based on the power factor curvature coefficient sequence, extract corresponding time indexes and compare them with turning points in the temperature change trend, calculate the time offset between the two, and generate a power factor-temperature inflection point offset sequence;

[0089] The locations with significant changes are identified from the power factor curvature coefficient sequence. The sliding extreme value detection method is used to extract local peak points. If the current point is higher than the two adjacent points and exceeds the significance threshold, it is recorded as the target point. The threshold is set based on the sequence mean and standard deviation. For example, if the mean is 0.015 and the standard deviation is 0.012, the threshold is set to 0.051. All time points that meet this criteria are extracted and formed into a peak time set. At the same time, the sign of the change rate of the temperature trend is detected. When the slope of two adjacent points changes from positive to negative or from negative to positive, it is a turning point. All such time points are extracted to form another set. Then, the time of each point in the two sets is compared, and the minimum time difference is taken as the offset value. For example, if the power factor changes significantly at 12, 24, and 39 seconds, and the temperature trend turning points are at 11, 26, and 42 seconds, the time offset is 1, 2, and 3 seconds. After calculating this offset for all points, an offset sequence is constructed for subsequent analysis.

[0090] S303: Based on the time interval relationship between the power factor-temperature inflection point offset sequence and the temperature change interference, determine whether there is a synchronous response phenomenon within the time period, extract the determination results and integrate the identification information to generate an environmental coupling interference determination identification sequence;

[0091] According to the time offset sequence between the above-mentioned power factor and temperature turning point, combined with the interference period information, a response judgment is made for each offset value, and the judgment threshold is set to 2 seconds. If the offset value corresponding to a certain time point does not exceed the threshold and is within the identified temperature interference period, it is judged that there is a response and recorded as 1, otherwise it is 0, forming a judgment result sequence. The interference period can be identified based on the external temperature source record or system status. For example, the interference period is defined as the 10th to 20th seconds and the 30th to 40th seconds. When the judgment point is at 12, 18, and 35 seconds, the corresponding offset values ​​are 1, 3, and 1 seconds. Among them, 12 and 35 seconds meet the threshold and interval conditions, and are marked as 1 in the output sequence. The 18-second offset exceeds the limit or is not within the interval range, and is marked as 0. Finally, a response identification sequence such as 1, 0, and 1 is generated.

[0092] The specific steps of S4 are:

[0093] S401: According to the environmental coupling interference determination identification sequence, the current density distribution and the voltage distortion spectrum signal in the unmarked time period are obtained, the associated data of the corresponding frequency points are extracted according to the frequency domain range, the voltage signal amplitudes at the frequency points are differenced and compared with the median of the amplitude distribution, the frequency points with the difference greater than the median are retained as frequency domain disturbance references, and the frequency domain disturbance reference frequency band interval values ​​are generated;

[0094] The current density and voltage distortion spectrum signals of the unmarked period are extracted from the power system operation data. After obtaining the signal, frequency domain analysis is required. Usually, the voltage amplitude information of multiple frequency points within the frequency range is obtained by transformation. For the amplitude data of each frequency point, difference processing is required to obtain the amplitude change between frequencies. For example, between 100Hz and 200Hz, if the voltage amplitude is 3.2 and 4.5, the difference is 1.3. By performing similar operations on all frequency points in the entire frequency range, a difference data set is formed, and then the data is collected from the set. Calculate the median. Assuming the difference set is 0.8, 1.1, 1.3, 1.5, and 2.2, the median is 1.3. Then compare the difference of each frequency point with the median, and only retain the frequency points with a difference greater than 1.3. For example, the difference between the corresponding frequencies of 250Hz and 350Hz is 1.5 and 2.2, which meets the retention conditions. Finally, these frequency points are used as frequency domain perturbation reference points, and the frequency band range is defined by the range before and after these points. If the reference frequency points are 250Hz and 350Hz, the reference frequency band range is 250Hz to 350Hz.

[0095] S402: Extracting the voltage signal amplitudes at the fundamental and third harmonic frequencies based on the frequency domain disturbance reference frequency band interval value, determining the energy proportion relationship between the two frequencies based on the spectrum amplitude distribution characteristics, and generating a fundamental and third harmonic energy ratio;

[0096] The voltage signal amplitudes at the fundamental and third harmonic frequencies are selected using the frequency domain perturbation reference frequency band interval value. If the reference frequency band is in the range of 200Hz to 400Hz, the voltage amplitudes corresponding to 50Hz and 150Hz need to be extracted. For example, during the measurement period, the amplitude at 50Hz is 4.0 and the amplitude at 150Hz is 1.5. Combined with the amplitude distribution trend of different frequency points in the spectrum graph, the energy ratio between the two frequency amplitudes is analyzed to determine the percentage of the third harmonic energy in the fundamental wave. If the fundamental amplitude is 4.0 and the third harmonic is 1.5, the third harmonic accounts for approximately 14%. This percentage value is associated with the distribution of abnormal frequency regions in the spectrum. In data analysis, this ratio is used to characterize the energy structure relationship between the main frequency signal and its harmonics, preparing for subsequent data feature integration.

[0097] S403: Combining the energy ratio of the fundamental wave to the third harmonic and the change in current density in the corresponding frequency band, constructing a joint vector reflecting the relationship between voltage and current changes, recording the frequency point mapping feature construction vector structure, calculating the joint vector construction value, and generating a calibration reference feature vector;

[0098] The specific calculation formula for calculating the joint vector construction value is:

[0099]

[0100] in, The mathematical symbol representing the voltage-current joint vector, E1 represents the energy measurement value of the fundamental component at the rated frequency, E3 represents the energy measurement value of the third harmonic component at 3 times the rated frequency, ΔJ represents the modulus of the current density variation in the corresponding frequency band, V k represents the instantaneous value of the voltage at the kth sampling point, Represents the current density vector of the kth frequency band, I k represents the instantaneous current value of the kth sampling point, γ represents the nonlinear regulation index of the instantaneous current value, and n represents the total number of sampling points;

[0101] The fundamental energy measurement value E1 = 5000J is derived from the measured data of the power quality analyzer at the rated frequency of 50Hz. The third harmonic energy measurement value E3 = 800J is derived from the monitoring data of the spectrum analyzer at the frequency range of 150Hz. The current density variation modulus ∥ΔJ∥ = 0.15A / m 2 The instantaneous voltage value V is obtained by collecting 100 sets of current density data within 0.1 seconds through the Hall sensor and calculating the standard deviation. k =235V is taken from the sampling value of the smart meter at t=kΔt, and the current density vector modulus is According to the conductor cross-sectional area 2.5mm 2 and the instantaneous value of current I k=9.8A is calculated by the formula J=I / S, the nonlinear adjustment index γ=0.8 is determined based on the fitting of the correlation curve between the current harmonic distortion rate and the system stability in the historical data, and the total number of sampling points n=1000 corresponds to the data volume of 1 second at a sampling frequency of 1 kHz.

[0102] The calculation process is carried out step by step:

[0103] Fundamental wave to harmonic energy ratio:

[0104]

[0105] The cube root term of the current density change:

[0106]

[0107] Single sampling point calculation item (taking k=1 as an example):

[0108]

[0109] Thousand-point sum approximation:

[0110]

[0111] The final joint vector modulus is:

[0112]

[0113] Parameter setting basis:

[0114] The E1 / E3 ratio range refers to the typical harmonic distribution of low-voltage power distribution systems in the IEEE 1459 standard;

[0115] The ∥ΔJ∥ value is derived from the allowable current density fluctuation threshold in the GB / T 17626-2017 electromagnetic compatibility test standard;

[0116] The γ value range of 0.5-1.2 is determined based on regression analysis of IEC 61000-4-15 flicker meter test data;

[0117] J k The cross-sectional area of ​​the conductor in the calculation is 2.5mm 2 Comply with GB / T3956 cable current carrying capacity specification;

[0118] The fluctuation range of voltage and current sampling values ​​meets the GB / T 12325 power quality supply voltage deviation standard.

[0119] The results show that the joint vector modulus reaches 1.336×10 8The magnitude reflects the strong coupling characteristics of the instantaneous voltage fluctuation and the current density change. The calculation results are directly used as the amplitude component of the calibration reference eigenvector to participate in the subsequent frequency mapping through the third harmonic energy weight correction and nonlinear scaling processing.

[0120] The specific steps of S5 are:

[0121] S501: Based on the calibration reference eigenvector, the pulse timestamps and corresponding harmonic frequency components in the metering pulse sequence of the electric energy meter are extracted. The phase difference interval of the pulse time sequence is determined according to the amplitude ratio of the frequency components in the eigenvector. The phase difference distribution interval coefficient is established based on the phase offset between the frequency components.

[0122] Based on the calibration reference feature vector, when extracting the pulse timestamps and corresponding harmonic frequency components in the metering pulse sequence, the rising edge of the TTL signal can be identified by edge detection on the voltage and current waveform signal with a sampling frequency of 100kHz, and the pulse timestamp can be extracted. At the same time, the waveform signal is analyzed in the frequency domain by fast Fourier transform to extract the amplitude and phase information of the main frequency, third harmonic, fifth harmonic, etc., and these frequency components are mapped to the amplitude ratio in the feature vector. For example, in a certain feature vector, the third harmonic accounts for 20% and the fifth harmonic accounts for 10%. The phase values ​​extracted from the two are 45 degrees and 90 degrees respectively. According to the frequency value, the time delay corresponding to the phase difference between them is about 0.0005 seconds. This delay can be used as a basic parameter that affects the distribution of pulse time between frequencies, and multiple phase difference intervals can be further delineated. For example, multiple adjacent intervals can be divided with a step size of 0.001 seconds. The number of pulse times corresponding to each frequency component is counted in each interval. These data are then normalized to form an interval coefficient matrix that describes the distribution of phase differences between frequencies. For example, the third harmonic has a high frequency in the interval [-0.001, 0], and the coefficient corresponding to this interval can be marked as 0.3. If the frequency is low in [-0.002, -0.001], the corresponding coefficient can be set to 0.1. The coefficients of all frequency components in all intervals are summarized to form a matching relationship between frequency and phase difference values, which is used to support the subsequent frequency area determination process.

[0123] S502: Matching the phase difference distribution interval coefficient with the harmonic energy ratio in the eigenvector, dividing the energy-dominant frequency interval according to the offset relationship, and measuring the output pitch difference of the interval according to the temporal position change of the pulse within the frequency interval to generate the output pitch change value;

[0124] When matching the phase difference distribution interval coefficients with the energy ratios of each frequency in the eigenvector, the energy of the frequency component must first be calculated based on the square of its amplitude. The energy of each frequency component is then normalized. For example, if the fundamental wave energy is 10,000, the third harmonic energy is 900, and the fifth harmonic energy is 400, the normalized proportions are approximately 68%, 18%, and 14%, respectively. These proportions are compared with the previously extracted interval coefficients. If the third harmonic accounts for a significantly larger proportion than other frequency components in a high-coefficient interval, the third harmonic can be determined to be the dominant energy frequency in that region and assigned to the frequency interval [100 Hz, 300 Hz]. Within this frequency interval, the temporal position distribution changes of the pulses are statistically analyzed. For example, if 100 pulses per second were detected in the early period and 110 pulses per second were detected more recently, the pitch is calculated to have changed from 10 milliseconds to approximately 9.1 milliseconds, a change of 0.9 milliseconds. This change is the output pitch change value, which is used as reference data for subsequent duty cycle adjustment.

[0125] S503: According to the correspondence between the output pitch variation value and the duty cycle setting value of the frequency component, the variation value is assigned to the original duty cycle sequence, the duty period of the output cycle is adjusted, and the multi-parameter fusion calibration instruction set is generated in combination with the time structure of the adjusted pulse sequence;

[0126] Matching processing is performed according to the output pitch change value and the duty cycle setting value corresponding to each frequency component. For example, the duty cycle of the fundamental wave is set to 50%, the third harmonic is 30%, and the fifth harmonic is 20%. When the output pitch change value is positive, such as 10 microseconds, it can be distributed according to the duty cycle ratio, which are approximately 6.8 microseconds, 1.8 microseconds, and 1.4 microseconds respectively. These incremental values ​​are added to the original duty period. For example, the original duty cycle of the fundamental wave is adjusted from 500 microseconds to 506.8 microseconds, and the remaining frequencies are also adjusted proportionally to form a new duty cycle sequence, further reconstructing the time structure of the entire pulse sequence. After adjustment, each pulse period contains multiple adjustment segments, each segment corresponds to a specific frequency component and has a new pulse width. Finally, an instruction set containing elements such as frequency, duty cycle increment, and time adjustment flag is integrated to be used for output control signals or data feedback instructions.

[0127] See also Figure 2 , a multi-parameter fusion smart electricity meter online calibration system, including:

[0128] The electric parameter co-sampling module collects voltage waveform sampling points, current harmonic components, and active power integral values, divides the time window according to the sampling sequence, extracts the three parameters within the time window, constructs a parameter matrix, extracts correlation features, and generates an electric parameter covariance feature set;

[0129] The anomaly marking module extracts the correlation change information between voltage and current based on the electrical parameter covariance feature set, compares it with the correlation change information of active power, and then matches and identifies the temperature change trend in the corresponding period to generate a multi-parameter anomaly correlation marking segment;

[0130] The curvature coupling module uses the multi-parameter anomaly association marker segment to obtain the power factor sequence within a specified period, identify the inflection point position during the change process, and simultaneously locate the change node in the temperature trend. It analyzes the positional relationship between the two types of feature points in the time series and generates an environmental coupling interference determination mark.

[0131] The frequency domain feature extraction module selects unmarked time period data based on the environmental coupling interference identification, extracts the current density sequence and voltage signal sequence, identifies the typical frequency components and energy characteristics of the two sets of signals, constructs a combined parameter vector, and generates a calibration reference feature vector.

[0132] The pulse compensation module extracts the metering pulse output interval and time parameters of the corresponding time period based on the calibration reference feature vector, identifies the correspondence between the pulse characteristics and the energy parameters, adjusts the pulse output rhythm parameters, and generates a multi-parameter fusion calibration instruction set.

[0133] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A multi-parameter fusion smart energy meter online calibration method, characterized in that: The following steps are involved: S1: Synchronously collect the voltage waveform sampling points, current harmonic components, and active power integral values ​​of the electric energy meter, divide them into time windows, calculate the covariance matrix of the parameters, and generate the electric parameter covariance feature set; S2: Based on the electrical parameter covariance feature set, extract the difference between the voltage and current covariance change rates and the active power covariance change rate. When the difference exceeds the covariance difference threshold, combine the change direction of the temperature sampling value to generate a multi-parameter anomaly association marker segment; S3: calling the multi-parameter anomaly association mark segment, performing curvature analysis on the power factor dynamic trajectory of the corresponding period, calculating the time series offset between the curvature extreme point and the temperature change inflection point, and generating an environmental coupling interference determination mark; S4: performing a joint frequency domain analysis of current density distribution and voltage distortion spectrum for an unmarked period according to the environmental coupling interference determination mark, extracting the fundamental wave and third harmonic energy ratio, and generating a calibration reference feature vector; S5: Based on the calibration reference eigenvector, perform phase synchronization compensation calculation on the electric energy meter measurement pulse sequence, adjust the duty cycle of the pulse output time interval according to the harmonic energy ratio in the eigenvector, and generate a multi-parameter fusion calibration instruction set.

2. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 1, characterized in that: The electrical parameter covariance feature set includes voltage covariance parameters, current covariance parameters, active power covariance parameters, and multi-parameter association strength indicators. The multi-parameter abnormal association mark segment includes a covariance difference out-of-limit mark, a temperature change direction mark, and an abnormal time period interval. The environmental coupling interference judgment mark includes the power factor curvature extreme value coordinates, the temperature inflection point timestamp, and the timing offset value. The calibration reference feature vector includes the fundamental wave energy ratio value, the third harmonic energy intensity value, and the frequency domain feature fusion coefficient. The multi-parameter fusion calibration instruction set includes a phase compensation parameter, a pulse duty cycle adjustment coefficient, and a harmonic energy calibration factor.

3. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 1, characterized in that: The specific steps of S1 are: S101: Collect voltage waveform sampling points, current harmonic components, and active power integral values, map the three types of data based on a unified time tag, divide them into time windows, construct parameter combinations under the windows, and generate multi-parameter synchronous set values; S102: using the parameters in the multi-parameter synchronization set value, extracting corresponding values ​​according to the time window, building a coordinated change relationship between the parameters based on the position relationship, and obtaining a window parameter covariance value group; S103: constructing a multi-dimensional feature representation based on the covariance values ​​between the parameters in the window parameter covariance quantity group, summarizing the structural sequence under the time window, and generating an electric parameter covariance feature set.

4. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 3 is characterized in that: The specific steps of S2 are: S201: Based on the electrical parameter covariance feature set, call the time series data of voltage, current, and active power, process the covariance change rate of voltage and current and the covariance change rate of active power within the time window, calculate the covariance dynamic adjustment factor, and aggregate the difference between the two to obtain an electrical parameter covariance change difference sequence; S202: Based on the electrical parameter covariance change difference sequence, calling the covariance difference threshold, identifying the time points exceeding the threshold in the difference sequence, and extracting the time index information in the original data to form an abnormality-related time index set; S203: For the abnormality-related time index set, call the temperature sampling value sequence of the corresponding time period, analyze the correlation between the temperature change direction and the electrical parameter change trend, and mark it as a multi-parameter abnormality-related marking segment.

5. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 4 is characterized in that: The specific calculation formula for calculating the covariance dynamic adjustment factor is: Where ΔΓ VI (t) represents the dynamic adjustment factor of voltage and current covariance at time t, ΔCov VI (t) represents the rate of change of the voltage and current covariance at time t, μ VI represents the sliding average of the voltage and current covariance within the time window, T represents the time window length constant, k represents the dynamic weight coefficient, σ VI (t) represents the standard deviation of the voltage and current covariance at time t, and the VI subscript specifically refers to the voltage and current parameter group.

6. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 4, characterized in that: The specific steps of S3 are: S301: Calling the time information in the multi-parameter anomaly associated mark segment, obtaining the corresponding power factor original trajectory and temperature change data, extracting the power factor change trend characteristics, calculating the curvature degree of the time point based on the trajectory structure change characteristics, and generating a power factor curvature coefficient sequence; S302: Identifying significant change locations based on the power factor curvature coefficient sequence, extracting corresponding time indexes and comparing them with turning points in the temperature change trend, calculating the time offset between the two, and generating a power factor-temperature inflection point offset sequence; S303: judging whether there is a synchronous response phenomenon within the time period based on the time interval relationship between the power factor-temperature inflection point offset sequence and the temperature change interference, extracting the judgment result and integrating the identification information to generate an environmental coupling interference judgment identification sequence.

7. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 6, characterized in that: The specific steps of S4 are: S401: According to the environmental coupling interference determination identification sequence, the current density distribution and the voltage distortion spectrum signal in the unmarked time period are obtained, the associated data of the corresponding frequency points are extracted according to the frequency domain range, the voltage signal amplitudes at the frequency points are differenced and compared with the median of the amplitude distribution, the frequency points with the difference greater than the median are retained as frequency domain disturbance references, and the frequency domain disturbance reference frequency band interval values ​​are generated; S402: Extracting the voltage signal amplitudes at the fundamental and third harmonic frequencies based on the frequency domain disturbance reference frequency band interval value, determining the energy proportion relationship between the two frequencies based on the spectrum amplitude distribution characteristics, and generating a fundamental and third harmonic energy ratio; S403: Combining the fundamental wave and third harmonic energy ratio with the current density change in the corresponding frequency band, constructing a joint vector reflecting the relationship between voltage and current changes, recording the frequency point mapping feature construction vector structure, calculating the joint vector construction value, and generating a calibration reference feature vector.

8. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 7, characterized in that: The specific calculation formula for calculating the joint vector construction value is: in, Represents the mathematical symbol of the voltage-current joint vector, E1 represents the energy measurement value of the fundamental component at the rated frequency, E3 represents the energy measurement value of the third harmonic component at 3 times the rated frequency, ΔJ represents the modulus of the current density change in the corresponding frequency band, V k represents the instantaneous value of the voltage at the kth sampling point, Represents the current density vector of the kth frequency band, I k represents the instantaneous current value at the kth sampling point, γ represents the nonlinear regulation index of the instantaneous current value, and n represents the total number of sampling points.

9. The multi-parameter fusion online calibration method for smart electric energy meters according to claim 7, characterized in that: The specific steps of S5 are: S501: Based on the calibration reference eigenvector, extract the pulse timestamps and corresponding harmonic frequency components in the metering pulse sequence of the electric energy meter, determine the phase difference interval of the pulse time sequence according to the amplitude ratio of the frequency components in the eigenvector, and establish the phase difference distribution interval coefficient based on the phase offset between the frequency components; S502: Matching the phase difference distribution interval coefficient with the harmonic energy ratio in the eigenvector, dividing the energy-dominant frequency interval according to the offset relationship, measuring the difference in output pitch of the interval according to the temporal position change of the pulse within the frequency interval, and generating an output pitch change value; S503: According to the correspondence between the output pitch variation value and the duty cycle setting value of the frequency component, the variation value is allocated to the original duty cycle sequence, the duty period of the output cycle is adjusted, and the multi-parameter fusion calibration instruction set is generated in combination with the time structure of the adjusted pulse sequence.

10. A multi-parameter fusion smart energy meter online calibration system, characterized in that: According to a multi-parameter fusion online calibration method for a smart electric energy meter according to any one of claims 1 to 9, the system comprises: The electric parameter co-sampling module collects voltage waveform sampling points, current harmonic components, and active power integral values, divides the time window according to the sampling sequence, extracts the three parameters within the time window, constructs a parameter matrix, extracts correlation features, and generates an electric parameter covariance feature set; The abnormal marking module extracts the correlation change information between voltage and current based on the electrical parameter covariance feature set, compares it with the correlation change information of active power, and then matches and identifies it in combination with the temperature change trend of the corresponding period to generate a multi-parameter abnormal correlation marking segment; The curvature coupling module calls the multi-parameter anomaly association marker segment to obtain the power factor sequence within a specified period, identify the inflection point position during the change process, and simultaneously locate the change node in the temperature trend. The positional relationship between the two types of feature points in the time series is analyzed to generate an environmental coupling interference determination mark; The frequency domain extraction module selects unmarked time period data according to the environmental coupling interference determination mark, extracts the current density sequence and the voltage signal sequence, identifies the typical frequency components and energy characteristics of the two sets of signals, constructs a combined parameter vector, and generates a calibration reference feature vector; The pulse compensation module extracts the metering pulse output interval and time parameters of the corresponding time period based on the calibration reference feature vector, identifies the correspondence between the pulse characteristics and the energy parameters, adjusts the pulse output rhythm parameters, and generates a multi-parameter fusion calibration instruction set.

Images