Electric energy metering method, system and device and storage medium

Abstract

The invention relates to the field of electric power, and discloses an electric energy metering method, system and device and a storage medium, and the method comprises the steps: employing an FIR zero compensation angle difference algorithm to correct a synchronization error angle difference between a current signal and a voltage signal, employing a self-convolution cascade zero-pole low-pass filter algorithm to carry out the filtering processing of a synchronization current signal and a synchronization voltage signal, and obtaining a current signal and a voltage signal; correcting a real-time phase difference between the filtering current signal and the filtering voltage signal by adopting a high-real-time FIR zero-step angular difference algorithm to obtain a corrected current signal and a corrected voltage signal, sampling to obtain current data and voltage data, and obtaining an initial electric energy metering result according to the current data and the voltage data; and obtaining environmental factors, calling a pre-trained electric energy metering error prediction model to obtain an electric energy metering prediction error, and obtaining a final electric energy metering result according to the initial electric energy metering result and the electric energy metering prediction error. The electric energy metering method with high accuracy and high linearity is provided, and the problem that an existing electric energy metering mode is large in measurement error is solved.

Description

Technical Field

[0001] The present invention belongs to the field of electric power and relates to an electric energy metering method, system, equipment and storage medium. Background Art

[0002] With the advancement of intelligent power grids and the increasing demand for precision in electricity management, the accuracy of electricity meters has become a key performance indicator. Traditional electricity meters are susceptible to factors such as temperature, humidity, and harmonics in complex environments, resulting in significant measurement errors. Existing calibration methods often fail to meet the high accuracy requirements of electricity meters across a wide current range. Summary of the Invention

[0003] The purpose of the present invention is to overcome the above-mentioned shortcomings of the prior art and provide an electric energy metering method, system, device and storage medium.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] In a first aspect, the present invention provides an electric energy metering method, comprising: obtaining a current signal and a voltage signal, and correcting the synchronization error angle difference between the current signal and the voltage signal using an FIR zero-compensation angular difference algorithm to obtain a synchronized current signal and a synchronized voltage signal; filtering the synchronized current signal and the synchronized voltage signal using a self-convolution cascaded zero-pole low-pass filter algorithm to obtain a filtered current signal and a filtered voltage signal; correcting the real-time phase difference between the filtered current signal and the filtered voltage signal using a high-real-time FIR zero-step angular difference algorithm to obtain a corrected current signal and a corrected voltage signal; sampling the corrected current signal and the corrected voltage signal to obtain current data and voltage data, and obtaining an initial electric energy metering result based on the current data and the voltage data; obtaining environmental factors and calling a pre-trained electric energy metering error prediction model to obtain an electric energy metering prediction error, and obtaining a final electric energy metering result based on the initial electric energy metering result and the electric energy metering prediction error.

[0006] Optionally, the adopting of the FIR zero compensation angle difference algorithm to correct the synchronization error angle difference between the current signal and the voltage signal includes: correcting the synchronization error angle difference between the current signal and the voltage signal by the following formula:

[0007]

[0008] Among them, I syn [n] is the synchronous current signal at time n; I[nk] is the current signal at time nk; U syn[n] is the synchronous voltage signal at time n; U[nk] is the voltage signal at time nk; h[k] is the impulse response coefficient of the filter; M is the order of the filter; the impulse response coefficient h[k] of the filter and the order M of the filter are determined according to the synchronization error angle difference between the current signal and the voltage signal.

[0009] Optionally, when the self-convolution cascade zero-pole low-pass filter algorithm is used to filter the synchronous current signal and the synchronous voltage signal, the low-pass filter in the self-convolution cascade zero-pole low-pass filter algorithm adopts a second-order elliptic filter, and the self-convolution operation adopts two self-convolutions of the transfer function of the second-order elliptic filter.

[0010] Optionally, the use of a high-real-time FIR zero-step angular difference algorithm to correct the real-time angular difference between the filtered current signal and the filtered voltage signal includes: obtaining a real-time phase difference Δφ(f) between the filtered current signal and the filtered voltage signal; using the filtered voltage signal as a corrected voltage signal, and obtaining a corrected current signal I′(f) by the following formula:

[0011] I′(f)=I(f)e -jΔφ(f)

[0012] Among them, e -jΔφ(f) is the phase compensation factor, and I(f) is the filtered current signal.

[0013] Optionally, the sampling and correcting the current signal and the correcting the voltage signal includes: adopting an analog-to-digital converter with a resolution of 24 bits or more to sample and correct the current signal and the correcting the voltage signal.

[0014] Optionally, the environmental factors include one or more of the following: temperature, humidity, air pressure, and external magnetic field strength.

[0015] Optionally, the pre-trained electric energy metering error prediction model is obtained by pre-training based on a least squares support vector machine model.

[0016] According to a second aspect of the present invention, an electric energy metering system is provided, comprising: a preprocessing module for acquiring a current signal and a voltage signal, and correcting the synchronization error angle difference between the current signal and the voltage signal using an FIR zero-compensation angular difference algorithm to obtain a synchronized current signal and a synchronized voltage signal; a filtering module for filtering the synchronized current signal and the synchronized voltage signal using a self-convolution cascaded zero-pole low-pass filter algorithm to obtain a filtered current signal and a filtered voltage signal; a correction module for correcting the real-time phase difference between the filtered current signal and the filtered voltage signal using a high-real-time FIR zero-step angular difference algorithm to obtain a corrected current signal and a corrected voltage signal; a calculation module for sampling the corrected current signal and the corrected voltage signal to obtain current data and voltage data, and obtaining an initial electric energy metering result based on the current data and voltage data; and a correction module for obtaining environmental factors and calling a pre-trained electric energy metering error prediction model to obtain an electric energy metering prediction error, and obtaining a final electric energy metering result based on the initial electric energy metering result and the electric energy metering prediction error.

[0017] According to a third aspect of the present invention, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned electric energy metering method when executing the computer program.

[0018] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned electric energy metering method are implemented.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The electric energy metering method of the present invention first uses an FIR zero-compensation angular difference algorithm to correct the angular difference and phase offset in the voltage and current signals, thereby ensuring synchronization between the voltage and current signals. A self-convolution cascaded zero-pole low-pass filter algorithm is then used to suppress noise and enhance low frequencies on the signal to obtain a high-quality, clean signal. During the real-time error correction phase, a high-real-time FIR zero-step angular difference algorithm is used to further optimize real-time performance and correct the real-time phase difference between the filtered current and filtered voltage signals, enabling the system to quickly respond to changes in current and voltage. Furthermore, a pre-trained electric energy metering error prediction model is invoked based on environmental factors to obtain the electric energy metering prediction error, which is used to correct the calculated initial electric energy metering result to obtain the final electric energy metering result. This method achieves high-accuracy and high-linearity electric energy metering, thereby effectively improving the measurement accuracy of the electric energy meter. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings of the embodiments of the present invention. Obviously, the drawings described below only relate to some embodiments of the present invention, and are not intended to limit the present invention.

[0022] Figure 1 This is a flow chart of an electric energy metering method according to an embodiment of the present invention.

[0023] Figure 2 Schematic diagram of the gain of the fourth-order self-convolution cascaded pole-zero low-pass filter at different frequencies according to an embodiment of the present invention.

[0024] Figure 3 Schematic diagram of the phase response of the fourth-order self-convolution cascaded pole-zero low-pass filter to different frequencies according to an embodiment of the present invention.

[0025] Figure 4 This is a signal waveform diagram before filtering by the fourth-order self-convolution cascaded zero-pole low-pass filter according to an embodiment of the present invention.

[0026] Figure 5 This is a waveform diagram of the signal after filtering by the fourth-order self-convolution cascade zero-pole low-pass filter according to an embodiment of the present invention.

[0027] Figure 6 This is a spectrum diagram of the signal before filtering by the fourth-order self-convolution cascade zero-pole low-pass filter according to an embodiment of the present invention.

[0028] Figure 7 This is a spectrum diagram of the signal after filtering by the fourth-order self-convolution cascade zero-pole low-pass filter according to an embodiment of the present invention.

[0029] Figure 8 This is a structural block diagram of an electric energy metering system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] The present invention is described in further detail below with reference to the accompanying drawings:

[0033] See also Figure 1 In one embodiment of the present invention, a method for electric energy metering is provided to solve the problem that traditional electric energy meters are easily affected by factors such as temperature, humidity and harmonics in complex environments, resulting in large measurement errors. This method realizes high-accuracy and high-linearity electric energy metering, effectively improving the measurement accuracy of the electric energy meter.

[0034] Specifically, the electric energy metering method of the present invention includes the following steps:

[0035] S1: Acquire current and voltage signals, and use the FIR zero compensation angle difference algorithm to correct the synchronization error angle difference between the current and voltage signals to obtain synchronized current and voltage signals.

[0036] S2: A self-convolution cascaded zero-pole low-pass filter algorithm is used to filter the synchronous current signal and the synchronous voltage signal to obtain a filtered current signal and a filtered voltage signal.

[0037] S3: A high real-time FIR zero-step angle difference algorithm is used to correct the real-time phase difference between the filtered current signal and the filtered voltage signal to obtain a corrected current signal and a corrected voltage signal.

[0038] S4: Sampling the corrected current signal and the corrected voltage signal to obtain current data and voltage data, and obtaining an initial electric energy measurement result based on the current data and voltage data.

[0039] S5: Acquire environmental factors and call a pre-trained electric energy metering error prediction model to obtain an electric energy metering prediction error, and obtain a final electric energy metering result based on the initial electric energy metering result and the electric energy metering prediction error.

[0040] The electric energy metering method of the present invention first uses an FIR zero-compensation angular difference algorithm to correct the angular difference and phase offset in the voltage and current signals, thereby ensuring synchronization between the voltage and current signals. A self-convolution cascaded zero-pole low-pass filter algorithm is then used to suppress noise and enhance low frequencies on the signal to obtain a high-quality, clean signal. During the real-time error correction phase, a high-real-time FIR zero-step angular difference algorithm is used to further optimize real-time performance and correct the real-time phase difference between the filtered current and filtered voltage signals, enabling the system to quickly respond to changes in current and voltage. Furthermore, a pre-trained electric energy metering error prediction model is invoked based on environmental factors to obtain the electric energy metering prediction error, which is used to correct the calculated initial electric energy metering result to obtain the final electric energy metering result. This method achieves high-accuracy and high-linearity electric energy metering, thereby effectively improving the measurement accuracy of the electric energy meter.

[0041] In a possible implementation, the adopting of the FIR zero compensation angle difference algorithm to correct the synchronization error angle difference between the current signal and the voltage signal includes: correcting the synchronization error angle difference between the current signal and the voltage signal by the following formula:

[0042]

[0043] Among them, I syn [n] is the synchronous current signal at time n; I[nk] is the current signal at time nk; U syn [n] is the synchronous voltage signal at time n; U[nk] is the voltage signal at time nk; h[k] is the impulse response coefficient of the filter; M is the order of the filter; the impulse response coefficient h[k] of the filter and the order M of the filter are determined according to the synchronization error angle difference between the current signal and the voltage signal.

[0044] Explanatory, highly sensitive current transformers (CTs) and voltage sensors are used to convert the current and voltage signals in the power grid into analog signals. High acquisition accuracy and a wide frequency response range are ensured to guarantee the integrity and accuracy of the original signal. In the process of electric energy metering, the phase synchronization of voltage and current signals is critical. If there is a phase offset or angular difference, the accuracy of electric energy metering will be affected. Using the FIR zero-compensation angular difference algorithm to correct the angular difference through digital filtering can ensure the synchronization of the signal current and voltage signals. Specifically, after signal acquisition, the voltage and current signals are first processed using the FIR zero-compensation angular difference algorithm to correct the angular difference (or phase offset) in the signal. Since the voltage and current signals may be out of sync, the FIR zero-compensation angular difference algorithm eliminates the angular difference problem caused by synchronization errors in the signal through precise digital filtering, ensuring that the voltage and current signals are always consistent, providing an accurate data basis for subsequent processing.

[0045] The core of the FIR zero-compensation angle error algorithm is to use the time-domain delay characteristics of the FIR filter to adjust the signal phase, synchronizing the voltage and current signals. Specifically, the FIR filter can control the signal delay by setting the filter coefficients and order, and use this delay to compensate for angle error and phase offset.

[0046] The specific implementation steps are as follows:

[0047] 1. Establish an FIR filter model. The basic form of the FIR filter is:

[0048]

[0049] Where y[n] is the filtered output signal, x[n] is the input signal (current or voltage), h[k] is the filter's impulse response coefficient, and M is the filter order. By adjusting the coefficient h[k], you can correct for phase shifts and ensure synchronization between the current and voltage signals.

[0050] 2. Determine the angular difference compensation amount. Before synchronization, it is necessary to measure the actual angular difference or phase offset between the current signal and the voltage signal, which is recorded as The angular difference can be obtained through system testing or real-time monitoring, and is usually expressed in the form of a time delay.

[0051] 3. Design a zero compensation filter. For the known angular difference The FIR filter is designed so that its delay characteristics can compensate for this angular difference. For example, if the angular difference is Δt, the filter's delay time should be approximately equal to Δt. By properly selecting the filter's order and coefficients, its frequency response and phase response can be synchronized, effectively compensating for the angular difference in the input signal.

[0052] 4. Apply filters for signal synchronization. Applying an FIR zero-compensation filter to the voltage and current signals synchronizes the current and voltage signals by adjusting the phase of the output signals according to the set coefficients, eliminating any angular differences and phase shifts. The filtered signals maintain consistent phase alignment, ensuring accurate results for energy calculations.

[0053] Exemplary, specific examples are provided for explanation:

[0054] The order is chosen to be third order for signal synchronization and compensation of the phase angle difference θ = 0.001° (corresponding to the time delay Δt = θ / (360·f), where f = 50 Hz is the power frequency). The output signal of the third-order FIR filter can be expressed as: y[n] = h[0]x[n] + h[1]x[n-1] + h[2]x[n-2] + h[3]x[n-3]. Where h[0], h[1], h[2], and h[3] are the impulse response coefficients of the filter, and the filter order is 3.

[0055] Compensation goal: Make the group delay of the signal close to the required time delay through the delay characteristics of the filter: Filtering target: at power frequency f c =50Hz maintains good low-frequency passability while attenuating high-frequency components. Coefficient design: Use Hanning window for design. The formula of Hanning window is: The impulse response of an ideal low-pass filter is: Among them, f c It is the cutoff frequency of the filter (expressed as normalized frequency), which determines the frequency range that the filter allows to pass. is the center point of the filter, and M+1 is the length of the filter (the filter order is M). The impulse response is infinitely long and needs to be truncated or weighted.

[0056] For this embodiment, the order is M=3, the filter length is M+1=4, and the center point By transforming the ideal impulse response h d [n] is multiplied by the window function W[n] to obtain the actual filter coefficients: h[0] = 0.112, h[1] = 0.276, h[2] = 0.276, h[3] = 0.112. The designed third-order FIR filter is applied to the current signal I[n] and the voltage signal U[n] to perform angle difference compensation processing: After filtering, I syn [n] and U syn The phase of [n] is synchronized, thus eliminating the angular difference between the original signals.

[0057] Through the above design and verification, the third-order FIR filter achieves accurate compensation of the diagonal difference, ensures the synchronization of voltage and current signals, and provides a guarantee for the accuracy of subsequent energy metering.

[0058] In one possible embodiment, when the self-convolution cascade zero-pole low-pass filter algorithm is used to filter the synchronous current signal and the synchronous voltage signal, the low-pass filter in the self-convolution cascade zero-pole low-pass filter algorithm adopts a second-order elliptic filter, and the self-convolution operation adopts two self-convolutions of the transfer function of the second-order elliptic filter.

[0059] Explanatory analysis shows that a self-convolution cascaded pole-zero low-pass filter algorithm is used to suppress noise and enhance low-frequency components in the signal, addressing the issue of non-integer sampling. This algorithm uses multiple stages of cascaded low-pass filters to suppress high-frequency noise in current and voltage signals while preserving the low-frequency components. This effectively eliminates the effects of external electromagnetic interference and equipment noise on the signal, resulting in more accurate and stable measurement data.

[0060] Explanatory, in order to reduce external noise interference and enhance the low-frequency components of the signal, a self-convolution cascade zero-pole low-pass filter algorithm is used. This algorithm consists of a cascade of multiple low-pass filters, and the zeros and poles of each filter are optimized by convolution, thereby effectively suppressing high-frequency noise.

[0061] In energy metering, collected voltage and current signals may contain high-frequency noise, which can affect the accuracy of energy calculations. The self-convolution cascaded pole-zero low-pass filter algorithm is a multi-order low-pass filtering method that progressively processes signals through multiple filtering stages, suppressing high-frequency noise and enhancing low-frequency components, achieving high-precision energy metering. In energy metering, the primary information in voltage and current signals is concentrated in the low-frequency range (typically around 50 Hz), while environmental factors, electromagnetic interference, and other factors can introduce high-frequency noise. Unfiltered noise can distort the current and voltage signals, further affecting accurate energy measurement. Therefore, suppressing noise and enhancing low-frequency signals during signal preprocessing are key steps in ensuring energy metering accuracy. The self-convolution cascaded pole-zero low-pass filter uses a series of filtering operations to gradually attenuate high-frequency noise while preserving the signal's low-frequency components. This filter relies primarily on pole-zero design, adjusting the filter's frequency response to effectively pass low-frequency signals while significantly attenuating high-frequency components. The design of zero poles can adjust the frequency response of the filter so that the filter has good pass characteristics for low frequency bands (low frequency enhancement) and strong attenuation for high frequency bands (noise suppression). The specific implementation steps are as follows:

[0062] 1. Filter self-convolution design. The concept of self-convolution is to re-input the output of the same filter as the input signal for a new round of filtering, thereby achieving a higher filter order. Specifically, if the original signal is filtered once to obtain the output signal y1(t), then y1(t) is used as the input for the next filtering, obtaining the output y2(t), and this process is repeated multiple times. This self-convolution operation makes the filtering effect smoother and more stable, suitable for suppressing high-frequency noise and improving the quality of low-frequency signals.

[0063] 2. Design and implementation of zero-pole filter. The transfer function of a typical low-pass filter can be expressed as: Where s is a complex frequency domain variable, ω c is the filter's cutoff frequency, and H(s) is the filter's transfer function. For cascaded filters, the overall expression for the transfer function is the product of the transfer functions of each individual filter: Among them, H i (s) is the transfer function of the i-th filter, and n is the number of filters in the cascade. In this embodiment, the transfer function of the low-pass filter can be expressed as: Where s is the complex frequency variable in Laplace transform; z1, z2, ..., z n It is the zero point of the filter, which determines the passability of the filter for a specific frequency band; p1, p2, ..., p m The zero and pole are the filter's poles, controlling the filter's attenuation of other frequency bands. By properly selecting the positions of the zero and pole, a filter can be constructed that has good passability in the low frequency band and significant attenuation in the high frequency band.

[0064] 3. Multi-order filter cascade. This algorithm is implemented through a multi-stage cascade approach. First, a low-pass filter H(s) is designed to perform preliminary filtering on the original signal. The filtered signal is then fed into a new filter, forming second-order, third-order, and even higher-order cascade filters. The cascade effect significantly improves the filter's ability to suppress high-frequency noise while maintaining the integrity of low-frequency signals. Multi-stage cascade filters further fine-tune the frequency response, achieving precise low-frequency enhancement and noise suppression.

[0065] 4. Signal output. After multi-stage filtering, the low-frequency components of the output voltage and current signals are retained, while high-frequency noise is fully suppressed, resulting in smooth and clear signals. This signal can be directly used for energy metering, ensuring high signal accuracy and low noise interference. The self-convolution cascaded zero-pole low-pass filter algorithm effectively suppresses high-frequency noise in the energy metering system, maintaining the integrity and clarity of the low-frequency signal. This enables the metering system to maintain high metering accuracy in actual operation, especially in complex power signal environments, thereby improving the overall performance of the energy meter.

[0066] In addition, in electric energy measurement, the sampling of current and voltage signals is usually completed by ADC (analog-to-digital converter). However, due to the sampling frequency f in practical applications, s The relationship between the signal frequency f0 and the signal frequency often cannot be accurately maintained as an integer multiple. This non-integer sampling may prevent high-frequency harmonic components from being completely canceled out by direct integration. In particular, the cos(2wt) component contained in the signal will leave residual high-frequency components within the integration interval. This residual component directly introduces measurement errors and affects the accuracy of the system.

[0067] To solve this problem, a self-convolution cascaded zero-pole low-pass filter algorithm is used. By performing multi-order low-pass filtering on the collected signal, the high-frequency residual components caused by non-integer sampling are gradually attenuated to a negligible level, while maintaining the integrity and accuracy of the low-frequency components. The specific implementation method is as follows:

[0068] 1. Mathematical analysis of the residual component. Assume that the original signal is: x(t) = A cos(wt) + B cos(2wt) + ..., where A cos(wt) is the main component (low frequency) and B cos(2wt) is the second harmonic component (high frequency). Under non-integer sampling, after integrating the second harmonic component cos(2wt), the theoretical integration result is: Since T is not necessarily an integer multiple of 2π / w, sin(2wt) will not completely cancel out between the upper and lower limits of the integral, resulting in a residue.

[0069] 2. The function of the filter. The principle of the self-convolution cascade zero-pole low-pass filter to weaken the cos(2wt) component is as follows: The transfer function of the filter is designed as: Among them, the zero point z i and the extreme point p i The position of is selected to strongly attenuate the high frequency signal 2w. By cascading multiple filters, the overall transfer function becomes: total (s)=H1(s)×H2(s)×…×H n (s), which makes the high-frequency attenuation effect exponentially enhanced.

[0070] 3. Filtering process. After the signal passes through the filter, the amplitude of the high-frequency component cos(2wt) is gradually reduced until it decays to a negligible level. Specifically, it is expressed as: y(t) = H total (s)·x(t), the filtered signal y(t) mainly retains the low-frequency component of cos(2wt), while the amplitudes of cos(2wt) and other high-frequency components are compressed to near zero.

[0071] For example, to meet the requirements of fully passing low-frequency signals below 20 Hz, significantly attenuating high-frequency signals, and minimizing response time, the following solution is adopted: Low-pass filter selection: elliptical filter (designed as a second-order filter). Self-convolution operation: convolve the filter output signal (the effect of cascading fourth-order filters).

[0072] 1. Elliptical filter design. Design goal: cutoff frequency f c =20Hz, sampling frequency f s Assuming 1000Hz, the passband ripple δ p =0.1dB, stopband attenuation δ s =40dB. The corresponding elliptic filter can be designed using MATLAB or Python's SciPy library: The calculated filter coefficients are: b = [0.0126, 0.0252, 0.0126], a = [1, -1.7786, 0.8008].

[0073] 2. Self-convolution (fourth-order filter implementation). The transfer function of the above elliptic filter is self-convolved twice to obtain the implementation equivalent to the fourth-order filter: the transfer function H(z) is: The self-convolution operation makes the transfer function become: The new coefficients are calculated as follows: numerator (b′): b′=b*b=[0.00015876,0.00063472,0.00095208,0.00063472,0.00015876]; denominator (a′): a′=a*a=[1,-3.5572,5.5175,-4.2772,1.6016].

[0074] 3. Filter Performance Analysis. Frequency Response: Passband: Below 20 Hz, the filter completely passes low-frequency signals, with signal amplitudes close to 1, ensuring the integrity of the low-frequency signal. Transition Band: Between 20 and 25 Hz, the filter amplitude rapidly decays, preventing high-frequency components from passing through. Stopband: Above 25 Hz, the filter effectively suppresses high-frequency components, with amplitude attenuation exceeding 40 dB. Response Time: The second-order elliptic filter has a fast response time. After two self-convolutions, the overall response time is two sampling periods, or approximately 2 ms. This meets the requirement for fast response. Residual Filtering Effect: After fourth-order filtering, high-frequency signals (such as the 50 Hz second harmonic) are effectively compressed, their amplitude attenuated to one millionth of the original signal, and the residual high-frequency effects are negligible. Through the design of the second-order elliptic filter and two self-convolution operations, a fourth-order low-pass filter is successfully implemented. It effectively passes low-frequency signals below 20 Hz, significantly suppresses high-frequency components above 25 Hz, and has a short response time (approximately 2 ms). This design is very suitable for electric energy measurement, especially when dealing with high-frequency residual problems caused by non-integer multiple sampling, which can significantly improve accuracy.

[0075] See also Figure 2 and Figure 3 , shows the frequency response of the fourth-order self-convolution cascaded zero-pole low-pass filter, showing the gain and phase response of the filter to different frequencies, verifying that signals below 20Hz can pass and high-frequency signals above 25Hz are effectively attenuated. Figure 4 and Figure 5 , shows the signal waveform of the fourth-order self-convolution cascade zero-pole low-pass filter, showing the signal before and after filtering. It can be clearly seen that the high-frequency components (such as 50Hz) are effectively suppressed. Figure 6 and Figure 7 , shows the spectrum diagram of the fourth-order self-convolution cascade zero-pole low-pass filter. The spectrum of the filtered signal should be close to zero in the high frequency band (such as 50Hz), while the low frequency band (such as 5Hz) remains basically unchanged.

[0076] In a possible implementation, the use of a high real-time FIR zero-step angle difference algorithm to correct the real-time angle difference between the filtered current signal and the filtered voltage signal includes:

[0077] Obtain the real-time phase difference Δφ(f) between the filtered current signal and the filtered voltage signal; use the filtered voltage signal as the corrected voltage signal and obtain the corrected current signal I′(f) using the following formula:

[0078] I′(f)=I(f)e -jΔφ(f)

[0079] Among them, e -jΔφ(f) is the phase compensation factor, and I(f) is the filtered current signal.

[0080] Explanatory, in order to ensure real-time performance, the present invention adopts a high real-time FIR zero-step angular difference algorithm. The algorithm can dynamically adjust the angular difference correction step size according to the rapid changes of current and voltage signals, improve the system response speed, and ensure that the electric energy metering system can make rapid adjustments when the load suddenly changes, avoiding metering errors caused by reaction lag. The algorithm adjusts the step size of the FIR filter by monitoring the changes of current and voltage in real time, so that the system can quickly respond to sudden changes in current and voltage. Mathematical formula of high real-time FIR zero-step angular difference algorithm: The core of the algorithm is to adjust the filter step size to adapt to real-time changes. Assuming that the frequency of the input signal changes greatly, the FIR zero-step angular difference algorithm adapts to this change by increasing or decreasing the step size Δt. For a standard FIR filter, the output of the input signal x[n] is: Where Δt is the step size adjustment, is the rate of change of the filter coefficients, which enables the filter step size to vary over time.

[0081] Explanatory note: In electric energy metering, real-time processing capabilities are crucial for coping with rapid changes in current and voltage. The high-real-time FIR zero-step angular difference algorithm can quickly respond to signal changes by optimizing the structure and calculation method of the FIR filter, thereby providing high-precision real-time correction during the metering process. In the actual working environment of the electric energy meter, the current and voltage signals may change rapidly due to load fluctuations or external interference. If the metering algorithm cannot respond quickly to these changes, it may lead to the accumulation of measurement errors, especially in a highly dynamic environment. Therefore, an algorithm that can quickly process signal changes is needed. FIR filters are widely used in the field of signal processing due to their linear phase and stability. However, the computational complexity of traditional FIR filters is high, and there may be a certain delay in the filtering process. In order to optimize real-time performance in electric energy metering, a high-real-time FIR zero-step angular difference algorithm is proposed. The core of this algorithm is to quickly correct the signal angular difference and phase offset by reducing the step update delay of the FIR filter and combining it with a zero compensation mechanism. The specific implementation steps are as follows:

[0082] 1. Optimization design of FIR filter. The output formula of traditional FIR filter is: Among them, y n is the filtered output at the current moment; h[k] is the coefficient of the FIR filter; x[nk] is the current and historical samples of the input signal.

[0083] 2. Zero-step angular error correction. Angular error is a measurement error caused by the phase asynchrony between the current and voltage signals. In traditional algorithms, angular error correction requires a long calculation time. However, in the high-real-time FIR algorithm, zero-step angular error correction is used to control the delay of signal synchronization within zero steps. Implementation methods include: directly estimating the signal phase difference Δφ(f): Δφ(f) = θ V (f)-θ I (f), where θ V (f) and θ I (f) are the phases of the voltage signal and the current signal at frequency f.

[0084] In order to correct the synchronization problem caused by the phase difference, the signal can be phase compensated in the frequency domain. Suppose the current signal i(t) needs to be phase corrected so that it can be synchronized with the voltage signal v(t). In the frequency domain, the phase of the signal can be adjusted by multiplying it by a phase compensation factor: I'(f) = I(f)e -jΔφ(f) , where Δφ(f) is the calculated phase difference, e -jΔφ(f) is the phase compensation factor. In this way, the current signal I′(f) is adjusted by the phase compensation factor in the frequency domain to synchronize it with the voltage signal.

[0085] 3. Real-time optimization. Parallel computing: split the operations of each level of the filter into parallel execution, and use the multi-core architecture of modern processors to accelerate the filtering calculation. Pipeline processing: through the pipeline architecture, complete the signal filtering, correction and output in stages to reduce the overall delay. This algorithm significantly improves the response speed of electric energy metering. In scenarios such as dynamic load switching and sudden current changes, it can track signal changes in real time and provide high-precision metering results. Test results show that the system's metering error is always kept within 0.02% throughout the entire current range (3mA ~ 1.2A), which is significantly better than traditional algorithms.

[0086] In a possible implementation manner, sampling and correcting the current signal and the voltage signal includes: using an analog-to-digital converter with a resolution of 24 bits or more to sample and correct the current signal and the voltage signal.

[0087] Explanatory, in the data acquisition phase, a high-performance ADC (analog-to-digital converter) is used to convert the optimized analog signal into a digital signal. The high-performance ADC has extremely low conversion error and fast sampling speed, which ensures high accuracy of the signal during the conversion process and avoids measurement deviations caused by sampling errors. The sampling rate of the ADC must be high enough to ensure accurate sampling of current and voltage signals. The mathematical formula of the ADC sampling process: For an analog signal x(t), its sampled value x[n] can be expressed by the following formula: Among them, Ts is the sampling period, and n is the index of the sampling point. A high-performance ADC has a precise quantization process, and its quantization error Δx can be expressed by the following formula: V max and V min are the maximum and minimum input voltages of the ADC, and N is the resolution of the ADC.

[0088] To illustrate, the implementation steps for a high-performance ADC in energy metering are as follows: 1. Optimize signal input. The ADC receives analog current and voltage signals that have been optimized by a pre-processing algorithm. The FIR zero-compensation angular error algorithm eliminates angular errors and phase offsets in the signals, ensuring signal synchronization. The self-convolution cascaded pole-zero low-pass filter algorithm removes high-frequency noise and enhances low-frequency signals, resulting in a purer signal. 2. High-precision digitization. Use a high-resolution (e.g., 24-bit) analog-to-digital converter (ADC) to ensure that critical signal details are not lost during the conversion process. Optimize the sampling rate to cover the required frequency range (e.g., 50 / 60 Hz) and provide sufficient redundancy. For example, within the full range of 3 mA to 1.2 A, the corresponding input voltage variation can be very small. A high-performance ADC can detect even the smallest changes in signal amplitude, ensuring high-precision measurement under both low and high current conditions. 3. Real-time data output. The digitized signal data is transmitted to a microprocessor or FPGA via a high-speed interface (e.g., SPI or I2C) for further processing by the LSSVM error prediction model and energy metering algorithm. The high-performance ADC ensures excellent metering performance over the entire measuring range of the electric energy meter through high-precision and high-speed signal conversion. It seamlessly collaborates with pre-algorithms (such as the FIR zero-compensation angle difference algorithm and the self-convolution cascade zero-pole low-pass filter algorithm) and post-models (such as the electric energy metering error prediction model), providing technical support for achieving high-linearity and high-accuracy electric energy metering.

[0089] Explanatory, the final electrical energy calculation can be done by the following formula: Where E is electric energy, U(t) and I(t) are voltage data and current data respectively, P(t) is instantaneous power, and T is the time interval.

[0090] In one possible implementation, the pre-trained electric energy metering error prediction model is pre-trained based on a least squares support vector machine model. In one possible implementation, the environmental factors include one or more of the following: temperature, humidity, air pressure, and external magnetic field strength.

[0091] The explanatory, energy metering error prediction model uses machine learning methods and environmental factors to predict and correct energy metering errors in real time. The model predicts energy metering errors based on real-time collected signals and environmental changes, and automatically adjusts compensation parameters to correct the errors.

[0092] Explanatory analysis shows that energy metering errors are not only related to signal characteristics such as current and voltage, but also closely related to external environmental factors such as temperature and humidity. For example, high temperatures can cause internal device parameter drift in the meter, affecting metering accuracy; high humidity can increase signal noise or cause changes in electrical characteristics.

[0093] The energy metering error prediction model is based on the LSSVM (least squares support vector machine) model. By training the LSSVM model, it can predict energy metering errors in real time and calibrate the energy metering system based on the prediction results. The LSSVM model uses current and voltage signals, as well as environmental factors (such as temperature and humidity) to model errors and perform regression analysis to reduce metering errors. Traditional error correction methods struggle to model these complex relationships in real time. However, the LSSVM model, with its nonlinear fitting capabilities and high computational efficiency, can effectively capture the relationship between energy metering errors and various influencing factors. Furthermore, the LSSVM model has strong generalization capabilities and can adapt to diverse environmental changes, thereby improving energy metering accuracy.

[0094] Mathematical formula of the energy metering error prediction model: The LSSVM model minimizes the error by solving the optimization problem. The objective function can be expressed as: Among them, α is the support vector coefficient, β is the coefficient of the regression function, C is the regularization parameter, ε i is the regression error, and N is the number of training samples. The regression function of the LSSVM model is usually: Among them, x is the input feature (including environmental factors, etc.), is a function mapped to a high-dimensional feature space, and β and b are the parameters of the regression function. By training and optimizing the LSSVM model, a model capable of predicting energy metering errors can be obtained, and the parameters used in the metering process can be adjusted in real time, thereby achieving high-precision energy metering.

[0095] The basic principle of the electric energy metering error prediction model based on the LSSVM model:

[0096] 1. The goal of the LSSVM model is to learn the mapping relationship between energy metering errors and input features (temperature, humidity, etc.), predict errors in real time, and make corrections. Its core goal is to solve the following optimization problem: Constraints: Among them, w is the weight vector; b is the bias; e iis the error; γ is the regularization parameter, which controls the trade-off between model complexity and error; For input data x i Through the above optimization objectives, LSSVM can find the optimal nonlinear mapping relationship between input features and measurement errors.

[0097] 2. Kernel function selection. The LSSVM model relies on the kernel function K(x, x′) to represent the nonlinear relationship of the input data. Typical choices include: Radial Basis Kernel (RBF kernel): The RBF kernel is able to capture complex nonlinear relationships and is the default choice for this model.

[0098] 3. Implementation steps of the electric energy metering error prediction model:

[0099] 1. Feature Extraction. The following features are collected in real time from the energy metering system: Environmental features: temperature T, humidity H. The input feature vector is: x = [T, H].

[0100] 2. Training Phase. Data Collection: Through experiments, we collect a large amount of sample data containing error information and record the true error y for each set of data. Model Training: Using the input data x and the true error y as the training set, we use the LSSVM algorithm to solve for the weights w and bias b to establish an error prediction model.

[0101] 3. Prediction and Correction. In actual operation, the real-time collected input data x is used to calculate the prediction error through the trained model: Among them, α i is the Lagrange multiplier, which is determined by the training phase. Finally, according to the prediction error Make real-time corrections to electric energy metering results.

[0102] Exemplary, specific examples are provided for explanation:

[0103] 1. Data example and normalization. Input feature range: Temperature T: [-45, 85] ℃; Humidity H: 10%, 85%; Normalization formula: Normalized eigenvector: x = [T norm ,H norm ].

[0104] 2. LSSVM kernel function parameters. Radial basis kernel function (RBF): Parameter setting: σ = 0.5; regularization parameter γ in the optimization objective: We set γ = 100 to balance the model complexity and prediction error.

[0105] 3. Specific parameters in the training phase. Through the training process, the support vector weight α iAssume the following three support vectors: x1 = [0.2, 0.1], x2 = [0.6, 0.4], x3 = [0.8, 0.7], with corresponding weights: α1 = 0.5, α2 = -0.2, α3 = 0.6. The bias is obtained by optimizing the problem: b = 0.01.

[0106] 4. Real-time prediction example. Input features: T = 250C, H = 50%. After normalization: x = [0.538, 0.533]. Kernel function operation: Prediction Error:

[0107] Assume that the original metering power value is: P 原始 =176W, then the corrected measurement value is:

[0108]

[0109] In the test environment, the energy calculation error prediction model combined with temperature and humidity compensation achieved significant results. Experimental results show that in extreme environments (temperature -45°C to 85°C, humidity 10% to 85%), the energy metering error after error correction is consistently maintained within 0.1%, significantly improving energy metering accuracy, enhancing the accuracy and stability of energy metering, and providing reliable energy metering support in various complex working environments.

[0110] Explanatory note: After undergoing optimization and processing at each of the aforementioned stages, the signal is input into the energy meter processing unit for accurate energy calculation and output. Practical application of the energy metering method of the present invention has demonstrated that even within the full current range of 3mA to 1.2A, the basic error in active power is maintained at less than 0.02%, requiring only a single-point calibration within the full current range. This method exhibits high accuracy and stability, making it widely applicable to fields such as power metering, load monitoring, and energy management.

[0111] The following are device embodiments of the present invention, which can be used to implement the method embodiments of the present invention. For details not disclosed in the device embodiments, please refer to the method embodiments of the present invention.

[0112] See also Figure 8 In another embodiment of the present invention, an electric energy metering system is provided, which can be used to implement the above-mentioned electric energy metering method. Specifically, the electric energy metering system includes a preprocessing module, a filtering module, a correction module, a calculation module and a calibration module.

[0113] Among them, the preprocessing module is used to obtain current signals and voltage signals, and use the FIR zero compensation angle difference algorithm to correct the synchronization error angle difference between the current signal and the voltage signal to obtain a synchronized current signal and a synchronized voltage signal; the filtering module is used to use the self-convolution cascade zero-pole low-pass filter algorithm to filter the synchronized current signal and the synchronized voltage signal to obtain a filtered current signal and a filtered voltage signal; the correction module is used to use the high real-time FIR zero-step angle difference algorithm to correct the real-time phase difference between the filtered current signal and the filtered voltage signal to obtain a corrected current signal and a corrected voltage signal; the calculation module is used to sample the corrected current signal and the corrected voltage signal to obtain current data and voltage data, and obtain the initial electric energy metering result based on the current data and voltage data; the correction module is used to obtain environmental factors and call the pre-trained electric energy metering error prediction model to obtain the electric energy metering prediction error, and obtain the final electric energy metering result based on the initial electric energy metering result and the electric energy metering prediction error.

[0114] All relevant contents of each step involved in the embodiment of the aforementioned electric energy metering method can be referred to the functional description of the corresponding functional modules of the electric energy metering system in the embodiment of the present invention, and will not be repeated here.

[0115] The module division in the embodiments of the present invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in various embodiments of the present invention may be integrated into a single processor, exist physically as separate modules, or two or more modules may be integrated into a single module. The integrated modules may be implemented in either hardware or software functional modules.

[0116] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the electric energy metering method.

[0117] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device, used to store programs and data. It is understood that the computer-readable storage medium herein may include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides storage space, which stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for being loaded and executed by a processor. These instructions may be one or more computer programs (including program code). It should be noted that the computer-readable storage medium herein may be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage device. The processor may load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the electric energy metering method in the above-mentioned embodiment.

[0118] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0119] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0120] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for measuring electric energy, characterized in that: include: Acquire current and voltage signals, and use FIR zero compensation angle difference algorithm to correct the synchronization error angle difference between the current and voltage signals to obtain synchronized current and voltage signals; The synchronous current signal and the synchronous voltage signal are filtered by using a self-convolution cascade zero-pole low-pass filter algorithm to obtain a filtered current signal and a filtered voltage signal; A high real-time FIR zero-step angle difference algorithm is used to correct the real-time phase difference between the filtered current signal and the filtered voltage signal to obtain a corrected current signal and a corrected voltage signal; Sampling the corrected current signal and the corrected voltage signal to obtain current data and voltage data, and obtaining an initial electric energy metering result based on the current data and voltage data; Environmental factors are obtained and a pre-trained electric energy metering error prediction model is called to obtain the electric energy metering prediction error, and a final electric energy metering result is obtained based on the initial electric energy metering result and the electric energy metering prediction error.

2. The electric energy metering method according to claim 1, characterized in that: The method of correcting the synchronization error angle difference between the current signal and the voltage signal by using the FIR zero compensation angle difference algorithm includes: The synchronization error angle difference between the current signal and the voltage signal is corrected by the following formula: Among them, I syn [n] is the synchronous current signal at time n; I[nk] is the current signal at time nk; U syn [n] is the synchronous voltage signal at time n; U[nk] is the voltage signal at time nk; h[k] is the impulse response coefficient of the filter; M is the order of the filter; the impulse response coefficient h[k] of the filter and the order M of the filter are determined according to the synchronization error angle difference between the current signal and the voltage signal.

3. The electric energy metering method according to claim 1, wherein: When the self-convolution cascade zero-pole low-pass filter algorithm is used to filter the synchronous current signal and the synchronous voltage signal, the low-pass filter in the self-convolution cascade zero-pole low-pass filter algorithm adopts a second-order elliptic filter, and the self-convolution operation adopts two self-convolutions of the transfer function of the second-order elliptic filter.

4. The electric energy metering method according to claim 1, wherein: The method of correcting the real-time angular difference between the filtered current signal and the filtered voltage signal by using a high-real-time FIR zero-step angular difference algorithm includes: Obtaining the real-time phase difference Δφ(f) between the filtered current signal and the filtered voltage signal; The filtered voltage signal is used as the corrected voltage signal, and the corrected current signal I′(f) is obtained by the following formula: I′(f)=I(f)e -jΔφ(f) Among them, e -jΔφ(f) is the phase compensation factor, and I(f) is the filtered current signal.

5. The electric energy metering method according to claim 1, characterized in that: The sampled corrected current signal and corrected voltage signal include: An analog-to-digital converter with a resolution of 24 bits or more is used to sample and correct the current signal and the voltage signal.

6. The electric energy metering method according to claim 1, characterized in that: The environmental factors include one or more of the following: temperature, humidity, air pressure, and external magnetic field strength.

7. The electric energy metering method according to claim 1, characterized in that: The pre-trained electric energy metering error prediction model is obtained by pre-training based on a least squares support vector machine model.

8. An electric energy metering system, characterized in that: include: The preprocessing module is used to obtain the current signal and the voltage signal, and use the FIR zero compensation angle difference algorithm to correct the synchronization error angle difference between the current signal and the voltage signal to obtain the synchronous current signal and the synchronous voltage signal; A filtering module is used to filter the synchronous current signal and the synchronous voltage signal using a self-convolution cascade zero-pole low-pass filter algorithm to obtain a filtered current signal and a filtered voltage signal; A correction module is used to correct the real-time phase difference between the filtered current signal and the filtered voltage signal using a high-real-time FIR zero-step angle difference algorithm to obtain a corrected current signal and a corrected voltage signal; A calculation module is used to sample the corrected current signal and the corrected voltage signal to obtain current data and voltage data, and obtain an initial electric energy measurement result based on the current data and voltage data; The correction module is used to obtain environmental factors and call a pre-trained electric energy metering error prediction model to obtain an electric energy metering prediction error, and obtain a final electric energy metering result based on the initial electric energy metering result and the electric energy metering prediction error.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the electric energy metering method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the electric energy metering method according to any one of claims 1 to 7 are implemented.

Images