Estimation Method and System, Device, and Storage Medium for On-Site Signal Flicker Parameters

Abstract

The present invention discloses a method and system, device, and storage medium for estimating the flicker parameters of on-site signals. The method for estimating the flicker parameters of on-site signals introduces a noise measurement matrix in the on-site signal acquisition process through an H∞ recursive estimator linear model, and combines the conversion between the estimated value of the state vector, the envelope, and the amplitude of the flicker component, so as to accurately estimate the amplitude values of multiple flicker components in the on-site signal, prevent being submerged by the measured noise, greatly improve the estimation accuracy of the flicker parameters of the on-site signal, and the conversion process is relatively simple, with high calculation efficiency, and can well meet the real-time requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy meter error testing, and in particular, to a method and system for estimating field signal flicker parameters, an electronic device, and a computer-readable storage medium. Background Art

[0002] From factory to on-site installation, electric energy meters must go through error testing and calibration procedures. During the daily operation of electric energy meters, due to force majeure factors, the measurement errors of electric energy meters may change. Especially in the situation of the rapid development of new energy and distributed power grids, with the continuous increase in the application of semiconductor loads, such as various switching power supplies, variable-frequency power supplies, speed regulators, LED lighting, high-power inverters and other dynamic loads, the load characteristics of the power grid become complex. Limited by the reproducibility of the virtual power test of the signal source, it is difficult for the electric energy meter measurement dynamic error model to fully simulate the unit response when the input measured current or power signal changes rapidly. At present, using on-site signals for real power simulation of electric energy meters has become an important means for electric energy meter error testing.

[0003] With the large-scale access of new energy grid connection and power electronic devices, power quality problems such as flicker in on-site collected signals are becoming increasingly serious, seriously affecting the accuracy of electric energy meter dynamic error testing. Therefore, how to accurately estimate the field signal flicker parameters is particularly important. At present, the commonly used method for estimating field signal flicker parameters is the method based on discrete Fourier transform. However, this method is greatly affected by noise and it is difficult to accurately analyze the modulation components in the flicker signal, resulting in large errors. In addition, there are currently methods for estimating flicker parameters using tools such as S transform and wavelet transform. Although the accuracy can be improved to a certain extent, the calculation process is very complex and the calculation efficiency is low, resulting in a reduction in the real-time performance of flicker parameter estimation. Summary of the Invention

[0004] The present invention provides a method and system for estimating field signal flicker parameters, an electronic device, and a computer-readable storage medium to solve the technical problem of large errors existing in the existing estimation of field signal flicker parameters based on discrete Fourier transform.

[0005] According to one aspect of the present invention, a method for estimating field signal flicker parameters is provided, including the following steps:

[0006] Collect an on-site signal sequence;

[0007] Construct a linear model of the H∞ recursive estimator for the on-site signal at time k, and set the initial iteration value;

[0008] Input the collected on-site signal sequence into the H∞ recursive estimator linear model for iterative operations to obtain the estimated value of the state vector of the on-site signal at time k;

[0009] Based on the estimated value of the state vector at time k, perform estimation calculations to obtain multiple envelopes at time k;

[0010] Input the multiple envelopes at time k into the trained adaptive linear perceptron neural network, and output the amplitude estimation values of multiple flicker components in the on-site signal.

[0011] Furthermore, the expression of the H∞ recursive estimator linear model is:

[0012] Z k =H k x + ω k

[0013] Where, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, ω k represents the noise measurement matrix during the on-site signal acquisition process, x represents the state vector of the on-site signal at time k,

[0014] N represents the number of harmonic components in the on-site signal at time k, represents the initial phase of the nth harmonic component in the on-site signal at time k, Envelope_n represents the envelope corresponding to the nth harmonic component of the on-site signal at time k, F represents the number of flicker components contained in the on-site signal at time k, V n represents the amplitude of the nth harmonic in the on-site signal, represents the relative amplitude or relative fluctuation of the ith flicker component, f fi and θ i respectively represent the frequency and phase angle of the ith flicker component, τ s represents the sampling period.

[0015] Furthermore, the estimated value of the state vector of the on-site signal at time k is specifically calculated based on the following formula:

[0016]

[0017] Where, represents the estimated value of the state vector at time k, φ represents the state transition matrix, which is a constant identity matrix, represents the estimated value of the state vector at time k - 1, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, Kk Denote the gain matrix as K k = P k-1 (I - αP k-1 +(H k ) T R -1 H k P k ) -1 φ T (H k ) T R -1 , P k denotes the covariance matrix of the estimation error at time k, P k-1 denotes the covariance matrix of the estimation error at time k - 1, I denotes the identity matrix, α denotes the error control factor, the superscript T denotes the matrix transpose, and R denotes the covariance of the noise measurement in the on-site signal acquisition process.

[0018] Furthermore, the covariance matrix P of the estimation error at time k k is updated in a recursive form during the estimation process, and the update expression is:

[0019] P k = φP k-1 (I - αP k-1 +(H k ) T R -1 H k P k-1 ) -1 φ T .

[0020] Furthermore, multiple envelopes at time k are obtained through estimation calculations based on the following formula:

[0021]

[0022] where denotes the estimated value of the envelope corresponding to the nth harmonic component of the on-site signal at time k, denotes the 2nth element in the estimated value of the state vector at time k, denotes the (2n + 1)th element in the estimated value of the state vector at time k.

[0023] Furthermore, the process of inputting the multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimation values of multiple flicker components in the on-site signal specifically includes the following content:

[0024] Take the N envelopes at time k as the input of the adaptive linear perceptron neural network, establish the weight vectors corresponding to the N envelopes at time k, determine the weight update method. After the weight of the adaptive linear perceptron neural network is updated, output the amplitude estimation value of the i-th flicker component in the on-site signal: represents the amplitude estimation value of the i-th flicker component in the on-site signal, represents the amplitude estimation value of the fundamental wave component in the on-site signal, represents the (2i + 1)-th element in the weight vector, represents the (2i + 2)-th element in the weight vector.

[0025] Furthermore, the weight update method is: w k+1 represents the weight vector at time k + 1, w k represents the weight vector at time k, e k represents the tracking error, λ represents an arbitrarily small value factor, the estimated value of the state vector at time k, the superscript T represents the matrix transpose, θ k represents the adaptive learning factor at time k, β is a constant, θ 0 represents the initial learning rate.

[0026] In addition, the present invention also provides an estimation system for the flicker parameters of the on-site signal, including:

[0027] A data acquisition module for collecting the on-site signal sequence;

[0028] A model construction module for constructing the H∞ recursive estimator linear model of the on-site signal at time k and setting the initial iteration value;

[0029] An iterative operation module for inputting the collected on-site signal sequence into the H∞ recursive estimator linear model for iterative operation to obtain the estimated value of the state vector of the on-site signal at time k;

[0030] An envelope calculation module for performing estimation calculations based on the estimated value of the state vector at time k to obtain multiple envelopes at time k;

[0031] An estimation analysis module for inputting the multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimation values of multiple flicker components in the on-site signal.

[0032] Furthermore, the H∞ recursive estimator linear model of the on-site signal at time k constructed by the model construction module is:

[0033] Z k = H k x + ωk

[0034] Among them, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, ω k represents the noise measurement matrix during the on-site signal acquisition process, and x represents the state vector of the on-site signal at time k,

[0035] N represents the number of harmonic components in the on-site signal at time k, represents the initial phase of the nth harmonic component in the on-site signal at time k, Envelope_n represents the envelope corresponding to the nth harmonic component of the on-site signal at time k, F represents the number of flicker components contained in the on-site signal at time k, V n represents the amplitude of the nth harmonic in the on-site signal, represents the relative amplitude or relative fluctuation of the ith flicker component, f fi and θ i respectively represent the frequency and phase angle of the ith flicker component, τ s represents the sampling period.

[0036] Furthermore, the iterative operation module specifically calculates the estimated value of the state vector of the on-site signal at time k based on the following formula:

[0037]

[0038] Among them, represents the estimated value of the state vector at time k, φ represents the state transition matrix, which is a constant identity matrix, represents the estimated value of the state vector at time k - 1, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, K k represents the gain matrix, K k = P k-1 (I - αP k-1 +(H k ) T R -1 H k P k ) -1 φ T (H k ) T R -1 , P k represents the covariance matrix of the estimation error at time k, P k-1represents the covariance matrix of the estimation error at time k-1, I represents the identity matrix, α represents the error control factor, the superscript T represents the matrix transpose, and R represents the covariance of the noise measurement in the on-site signal acquisition process.

[0039] Further, the covariance matrix P of the estimation error at time k k is updated in a recursive form during the estimation process, and the update expression is:

[0040] P k = φP k-1 (I - αP k-1 +(H k ) T R -1 H k P k-1 ) -1 φ T 。

[0041] Further, the envelope calculation module performs estimation calculations based on the following formula to obtain multiple envelopes at time k:

[0042]

[0043] Where, represents the estimated value of the envelope corresponding to the nth harmonic component of the on-site signal at time k, represents the 2nth element in the estimated value of the state vector at time k, represents the (2n + 1)th element in the estimated value of the state vector at time k.

[0044] Further, the process of the estimation analysis module inputting multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimation values of multiple flicker components in the on-site signal specifically includes the following content:

[0045] Taking the N envelopes at time k as the input of the adaptive linear perceptron neural network, establishing the weight vector corresponding to the N envelopes at time k, determining the method for updating the weights. After the weights of the adaptive linear perceptron neural network are updated, the amplitude estimation value of the ith flicker component in the on-site signal is output: represents the amplitude estimation value of the ith flicker component in the on-site signal, represents the amplitude estimation value of the fundamental wave component in the on-site signal, represents the (2i + 1)th element in the weight vector, represents the (2i + 2)th element in the weight vector.

[0046] Further, the method for updating the weights is: w k+1denotes the weight vector at time k+1, w k denotes the weight vector at time k, e k denotes the tracking error, and λ denotes an arbitrarily small value factor the estimated value of the state vector at time k, and the superscript T denotes the matrix transpose, θ k denotes the adaptive learning factor at time k β is a constant, θ 0 denotes the initial learning rate

[0047] In addition, the present invention also provides an electronic device, including a processor and a memory. A computer program is stored in the memory, and the processor is configured to execute the steps of the method described above by calling the computer program stored in the memory

[0048] In addition, the present invention also provides a computer-readable storage medium for storing a computer program for estimating the flicker parameters of on-site signals. When the computer program runs on a computer, it executes the steps of the method described above

[0049] The present invention has the following effects

[0050] For the method for estimating the flicker parameters of on-site signals according to the present invention, first, a linear model of the H∞ recursive estimator for the on-site signal at time k is constructed, and a noise measurement matrix in the process of collecting the on-site signal is introduced. Then, the collected on-site signal sequence is input into the linear model of the H∞ recursive estimator for iterative operation, and the estimated value of the state vector of the on-site signal at time k can be obtained. Based on the estimated value of the state vector at time k, multiple envelopes at time k are calculated through estimation. Finally, the multiple envelopes at time k are used as the input of the adaptive linear perceptron neural network, and the amplitude estimation values of multiple flicker components in the on-site signal can be output through the weight update of the adaptive linear perceptron neural network. By introducing the noise measurement matrix in the process of collecting the on-site signal through the linear model of the H∞ recursive estimator and combining the conversion between the estimated value of the state vector, the envelope, and the amplitude of the flicker component, the amplitude values of multiple flicker components in the on-site signal can be accurately estimated, preventing being submerged by the measured noise, greatly improving the estimation accuracy of the flicker parameters of the on-site signal, and the conversion process is relatively simple with high calculation efficiency, which can well meet the real-time requirement

[0051] In addition, the system for estimating the flicker parameters of on-site signals according to the present invention also has the above advantages

[0052] In addition to the objectives, features, and advantages described above, the present invention has other objectives, features, and advantages. The present invention will be further described in detail below with reference to the drawings Description of the Drawings

[0053] The accompanying drawings, which form a part of this application, are used to provide a further understanding of the present invention. The schematic embodiments and descriptions thereof of the present invention are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0054] Figure 1 is a schematic flow chart of a method for estimating the on-site signal flicker parameters of a preferred embodiment of the present invention.

[0055] Figure 2 is a schematic diagram of performing an estimation operation using an adaptive linear perceptron neural network in a preferred embodiment of the present invention.

[0056] Figure 3 is a schematic diagram of the module structure of an on-site signal flicker parameter estimation system according to another embodiment of the present invention. Detailed Embodiment

[0057] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, the present invention can be implemented in many different ways defined and covered by the following.

[0058] As Figure 1 shown, a preferred embodiment of the present invention provides a method for estimating on-site signal flicker parameters, including the following:

[0059] Step S1: Collect the on-site signal sequence;

[0060] Step S2: Construct a linear model of the H∞ recursive estimator for the on-site signal at time k and set the initial iteration value;

[0061] Step S3: Input the collected on-site signal sequence into the linear model of the H∞ recursive estimator for iterative operation to obtain the estimated value of the state vector of the on-site signal at time k;

[0062] Step S4: Perform an estimation calculation based on the estimated value of the state vector at time k to obtain multiple envelopes at time k;

[0063] Step S5: Input the multiple envelopes at time k into the trained adaptive linear perceptron neural network and output the amplitude estimated values of multiple flicker components in the on-site signal.

[0064] It can be understood that for the method of estimating the flicker parameters of on-site signals in this embodiment, first, a linear model of the H∞ recursive estimator for the on-site signal at time k is constructed, and a noise measurement matrix in the on-site signal acquisition process is introduced. Then, the acquired on-site signal sequence is input into the linear model of the H∞ recursive estimator for iterative calculation, and the estimated value of the state vector of the on-site signal at time k can be obtained. Based on the estimated value of the state vector at time k, estimation calculations are performed to obtain multiple envelopes at time k. Finally, the multiple envelopes at time k are used as the input to the adaptive linear perceptron neural network, and the amplitude estimation values of multiple flicker components in the on-site signal can be output through the weight update of the adaptive linear perceptron neural network. By introducing the noise measurement matrix in the on-site signal acquisition process through the H∞ recursive estimator linear model and combining the conversion between the estimated value of the state vector, the envelope, and the amplitude of the flicker component, the amplitude values of multiple flicker components in the on-site signal can be accurately estimated, preventing being submerged by the measured noise, greatly improving the estimation accuracy of the flicker parameters of the on-site signal. Moreover, the conversion process is relatively simple and the calculation efficiency is high, which can well meet the real-time requirements.

[0065] It can be understood that in step S1, an on-site signal sequence is collected and denoted as Z k ={Z 1 , Z 2 ,..., Z k , Z k +1 ,..., Z K}, where K represents the number of on-site signals collected. In addition, the on-site test signal can be generated by the three-phase standard power source Fluke6100 according to the IEC61000-4-15 standard. The test signal at time k can be defined as:

[0066] Z k =A1{Envelope_1}+A2{Envelope_2}+...+A n {Envelope_n}+...+A N {Envelope_N}+σrandn k

[0067] where randn k represents additive Gaussian noise, σ represents the noise standard deviation, Envelope_n represents the envelope corresponding to the nth harmonic component of the on-site signal at time k, N represents the number of harmonic components included in the on-site signal, and A n represents the nth harmonic component. Based on trigonometric function decomposition, A n can be obtained: f represents the fundamental frequency of the on-site signal at time k, and τ s represents the sampling period. Represents the initial phase of the nth harmonic component in the on-site signal.

[0068] It can be understood that in the step S2, first construct the H∞ recursive estimator linear model:

[0069] Z k =H k x + ω k

[0070] Where, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, ω k represents the noise measurement matrix during the acquisition process of the on-site signal, and x represents the state vector of the on-site signal at time k.

[0071] N represents the number of harmonic components in the on-site signal at time k. represents the initial phase of the nth harmonic component in the on-site signal at time k, Envelope_n represents the envelope corresponding to the nth harmonic component of the on-site signal at time k, and the expression is: F represents the number of flicker components contained in the on-site signal at time k, V n represents the amplitude of the nth harmonic in the on-site signal. represents the relative amplitude or relative fluctuation of the ith flicker component, where V t represents the effective value of the signal, ΔV i represents the amplitude of the ith flicker component, f fi and θ i respectively represent the frequency and phase angle of the ith flicker component, and τ s represents the sampling period.

[0072] Then, set the value at time 0, that is, the initial iteration value. Specifically: the covariance matrix P of the estimation error at time 0 0 is a matrix of dimension 74*74, and the element values are all 103. The estimated value of the state vector at time 0 is a 1*74 vector, and its element values are all 1.

[0073] It can be understood that since the amplitude value of the flicker component in the on-site signal is small, it is easily submerged by the measurement noise, so that the amplitude value of the flicker component cannot be accurately estimated. And in the present invention, by constructing the H∞ recursive estimator linear model of the on-site signal at time k, the noise measurement matrix ω during the acquisition process of the on-site signal kIntroduced, during subsequent iterative operations, the measurement noise and the state vector can be separately characterized and operated on, which is conducive to accurately extracting the amplitude value of the flicker component in the on-site signal, greatly improving the estimation accuracy.

[0074] It can be understood that in step S3, the collected on-site signal sequence Z k ={Z 1 , Z 2 ,..., Z k , Z k+1 ,..., Z K} is input into the H∞ recursive estimator linear model for iterative operations. In the time domain, the state vector x is estimated by the measured value Z k and the following time update equation. The specific expression of the time update equation is:

[0075] Among them, represents the estimated value of the state vector at time k, φ represents the state transition matrix, which is a constant identity matrix, represents the estimated value of the state vector at time k - 1, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, K k represents the gain matrix, K k = P k-1 (I - αP k-1 +(H k ) T R -1 H k P k ) -1 φ T (H k ) T R -1 , P k represents the covariance matrix of the estimation error at time k, P k-1 represents the covariance matrix of the estimation error at time k - 1, I represents the identity matrix, α represents the error control factor, the superscript T represents the matrix transpose, and R represents the covariance of the noise measurement during the on-site signal acquisition.

[0076] Optionally, to minimize the estimation error in the worst case, the covariance matrix P k of the estimation error at time k is updated in a recursive form during the estimation process. The update expression is:

[0077] P k = φP k-1 (I - αP k-1 +(H k ) T R-1 H k P k-1 ) -1 φ T 。

[0078] It can be understood that in the step S4, the estimation calculation is specifically performed based on the following formula to obtain multiple envelopes at the k-th moment:

[0079]

[0080] where, represents the estimated value of the envelope corresponding to the n-th harmonic component of the on-site signal at the k-th moment, represents the 2n-th element in the estimated value of the state vector at the k-th moment, represents the (2n + 1)-th element in the estimated value of the state vector at the k-th moment.

[0081] It can be understood that in the step S5, the process of inputting multiple envelopes at the k-th moment into the trained adaptive linear perceptron neural network and outputting the estimated values of the amplitudes of multiple flicker components in the on-site signal specifically includes the following content:

[0082] Taking the N envelopes at the k-th moment as the input of the adaptive linear perceptron neural network, and establishing the weight vector w corresponding to the N envelopes at the k-th moment k , Determining the method for updating the weights. After the adaptive linear perceptron neural network updates the weights, it outputs the estimated value of the amplitude of the i-th flicker component in the on-site signal: represents the estimated value of the amplitude of the i-th flicker component in the on-site signal, represents the estimated value of the amplitude of the fundamental wave component in the on-site signal, represents the (2i + 1)-th element in the weight vector, represents the (2i + 2)-th element in the weight vector.

[0083] It can be understood that as Figure 2 shown, the input of the adaptive linear perceptron neural network includes 74 elements, and the initialization vector of the 74 input elements is: In addition, the initial weight vector w 0 is a 1*74 vector, and the element values are all 1.

[0084] In addition, according to the Widrow-Hoff criterion, the method for updating the weights is set as: w k+1 represents the weight vector at the (k + 1)-th moment, w k represents the weight vector at the k-th moment, e krepresents the tracking error, and λ represents an arbitrarily small value factor used to avoid the denominator being zero. The estimated value of the state vector at time k, where the superscript T represents the matrix transpose, and θ k represents the adaptive learning factor at time k. β is a constant, and θ 0 represents the initial learning rate.

[0085] It can be understood that in an embodiment of the present invention, according to the IEC 61000-4-15 standard, the voltage signal contains 36 significant flicker components, and the following test signals are generated using a three-phase standard power source Fluke 6100:

[0086]

[0087] where τ s = 1200, F = 36, f = 50Hz, and σ takes the value of 0.02.

[0088] Then, the generated test signals are collected on-site, and the on-site signal at the first observation time, that is, at time k = 1, is denoted as Z 1 to obtain the on-site signal sequence Z k .

[0089] Then, an H∞ recursive estimator linear model is established, and the values at time 0, that is, the initial iteration values, are set as follows: The covariance matrix P of the estimation error at time 0 0 is a matrix of dimension 74*74, and all element values are 10 3 , is a 1*74 vector, and all element values are 1. The estimated value of the state vector x at time k is estimated using the formula where α = 8 and R = 0.007. As the sampling value k of the on-site signal changes, the estimated value of the state vector x is continuously iterated to obtain the estimated value of the state vector at time k Furthermore, based on the formula: the N envelopes at time k are estimated.

[0090] The N envelopes obtained are used as the input of the adaptive linear perceptron neural network, the weight vector of the N envelopes at time k is constructed, the method for updating the weights is determined, and λ = 0.0001 and β = 1 are taken, and θ 0 = 5. After the adaptive linear perceptron neural network performs the weight update calculation, the amplitude estimation value of the i-th flicker component in the on-site signal is output: After 14 fundamental wave periods, the flicker amplitude estimation value has converged. Among them, the specific estimation results are shown in Table 1.

[0091] Table 1. Flicker frequency f of the signal fi (Hz) and the relative amplitude estimation results of the corresponding flicker components

[0092]

[0093] In addition, as Figure 3 shown, another embodiment of the present invention further provides an estimation system for the flicker parameters of on-site signals. Preferably, the above-mentioned estimation method is adopted. The system includes:

[0094] A data acquisition module for acquiring on-site signal sequences;

[0095] A model construction module for constructing a linear model of the H∞ recursive estimator of the on-site signal at time k and setting the initial iteration value;

[0096] An iterative operation module for inputting the acquired on-site signal sequence into the linear model of the H∞ recursive estimator for iterative operation to obtain the estimated value of the state vector of the on-site signal at time k;

[0097] An envelope calculation module for performing estimation calculations based on the estimated value of the state vector at time k to obtain multiple envelopes at time k;

[0098] An estimation analysis module for inputting multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimation values of multiple flicker components in the on-site signal.

[0099] It can be understood that for the estimation system of the on-site signal flicker parameters in this embodiment, first, a linear model of the H∞ recursive estimator of the on-site signal at time k is constructed, and the noise measurement matrix in the on-site signal acquisition process is introduced. Then, the acquired on-site signal sequence is input into the linear model of the H∞ recursive estimator for iterative operation, and the estimated value of the state vector of the on-site signal at time k can be obtained. Based on the estimated value of the state vector at time k, estimation calculations are performed to obtain multiple envelopes at time k. Finally, multiple envelopes at time k are used as the input of the adaptive linear perceptron neural network, and the amplitude estimation values of multiple flicker components in the on-site signal can be output through the weight update of the adaptive linear perceptron neural network. The present invention introduces the noise measurement matrix in the on-site signal acquisition process through the linear model of the H∞ recursive estimator, and combines the conversion between the estimated value of the state vector, the envelope, and the amplitude of the flicker component, so as to accurately estimate the amplitude values of multiple flicker components in the on-site signal, prevent being submerged by the measured noise, greatly improve the estimation accuracy of the on-site signal flicker parameters, and the conversion process is relatively simple with high calculation efficiency, which can well meet the real-time requirements.

[0100] It can be understood that the linear model of the H∞ recursive estimator of the on-site signal at time k constructed by the model construction module is:

[0101] Z k = H k x + ω k

[0102] Wherein, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, ω k represents the noise measurement matrix during the on-site signal acquisition process, and x represents the state vector of the on-site signal at time k,

[0103] N represents the number of harmonic components in the on-site signal at time k, represents the initial phase of the nth harmonic component in the on-site signal at time k, Envelope_n represents the envelope corresponding to the nth harmonic component of the on-site signal at time k, F represents the number of flicker components included in the on-site signal at time k, V n represents the amplitude of the nth harmonic in the on-site signal, represents the relative amplitude or relative fluctuation of the ith flicker component, f fi and θ i respectively represent the frequency and phase angle of the ith flicker component, τ s represents the sampling period. It can be understood that the iterative operation module specifically calculates the estimated value of the state vector of the on-site signal at time k based on the following formula:

[0104]

[0105] Wherein, represents the estimated value of the state vector at time k, φ represents the state transition matrix, which is a constant identity matrix, represents the estimated value of the state vector at time k - 1, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, K k represents the gain matrix, K k = P k-1 (I - αP k-1 +(H k ) T R -1 H k P k ) -1 φ T (H k ) T R -1 , P k represents the covariance matrix of the estimation error at time k, P k-1denotes the covariance matrix of the estimation error at time k - 1, I denotes the identity matrix, α denotes the error control factor, the superscript T denotes the matrix transpose, and R denotes the covariance of the noise measurement during the on - site signal acquisition.

[0106] Among them, the covariance matrix P of the estimation error at time k k is updated in a recursive form during the estimation process, and the update expression is:

[0107] P k = φP k-1 (I - αP k-1 +(H k ) T R -1 H k P k-1 ) -1 φ T .

[0108] It can be understood that the envelope calculation module performs estimation calculations based on the following formula to obtain multiple envelopes at time k:

[0109]

[0110] Among them, denotes the estimated value of the envelope corresponding to the nth harmonic component of the on - site signal at time k, denotes the 2nth element in the estimated value of the state vector at time k, denotes the (2n + 1)th element in the estimated value of the state vector at time k.

[0111] It can be understood that the process of the estimation and analysis module inputting multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimation values of multiple flicker components in the on - site signal specifically includes the following contents:

[0112] Taking the N envelopes at time k as the input of the adaptive linear perceptron neural network, establishing the weight vectors corresponding to the N envelopes at time k, determining the method for updating the weights. After the adaptive linear perceptron neural network updates the weights, it outputs the amplitude estimation value of the ith flicker component in the on - site signal: denotes the amplitude estimation value of the ith flicker component in the on - site signal, denotes the amplitude estimation value of the fundamental wave component in the on - site signal, denotes the (2i + 1)th element in the weight vector, denotes the (2i + 2)th element in the weight vector.

[0113] Among them, the method for updating the weights is: w k+1denotes the weight vector at time k + 1, w k denotes the weight vector at time k, e k denotes the tracking error, and λ denotes an arbitrarily small value factor the estimated value of the state vector at time k, and the superscript T represents the matrix transpose, θ k denotes the adaptive learning factor at time k β is a constant, θ 0 denotes the initial learning rate

[0114] It can be understood that each module in the system of this embodiment corresponds to each step of the above method embodiment respectively. Therefore, the specific working principle and working process of each module can be referred to the above method embodiment, and will not be elaborated here

[0115] In addition, another embodiment of the present invention further provides an electronic device, including a processor and a memory. A computer program is stored in the memory, and the processor is used to execute the steps of the above method by calling the computer program stored in the memory

[0116] In addition, another embodiment of the present invention further provides a computer-readable storage medium for storing a computer program for estimating the flicker parameters of on-site signals. When the computer program runs on a computer, it executes the steps of the above method

[0117] The forms of common computer-readable storage media generally include: floppy disk, flexible disk, hard disk, magnetic tape, any other magnetic medium, CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with a pattern of holes, random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), flash erasable programmable read-only memory (FLASH-EPROM), any other memory chip or cartridge, or any other medium that can be read by a computer. The instructions can be further transmitted or received by a transmission medium. The term transmission medium can include any tangible or intangible medium that can be used to store, encode, or carry instructions for execution by a machine, and includes digital or analog communication signals or other intangible media that facilitate the communication of the above instructions. The transmission medium includes coaxial cables, copper wires, and optical fibers, which include the wires of a bus used to transmit a computer data signal

[0118] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, various modifications and variations can be made to the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An estimation method for the flicker parameters of on-site signals, characterized in that, It includes the following: Collect the on-site signal sequence; Construct the linear model of the H∞ recursive estimator for the on-site signal at time k, and set the initial iteration value; Input the collected on-site signal sequence into the linear model of the H∞ recursive estimator for iterative operation, and obtain the estimated value of the state vector of the on-site signal at time k; Perform estimation calculations based on the estimated value of the state vector at time k to obtain multiple envelopes at time k; Input the multiple envelopes at time k into the trained adaptive linear perceptron neural network, and output the amplitude estimated values of multiple flicker components in the on-site signal.

2. The method for estimating the on-site signal flicker parameter according to claim 1, wherein The expression of the linear model of the H∞ recursive estimator is: Z k = H k x + ω k Among them, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, ω k represents the noise measurement matrix during the on-site signal acquisition process, x represents the state vector of the on-site signal at time k, N represents the number of harmonic components in the on-site signal at time k, represents the initial phase of the nth harmonic component in the on-site signal at time k, Envelope_n represents the envelope corresponding to the nth harmonic component of the on-site signal at time k, F represents the number of flicker components contained in the on-site signal at time k, V n represents the amplitude of the nth harmonic in the on-site signal, represents the relative amplitude or relative fluctuation of the ith flicker component, f fi and θ i respectively represent the frequency and phase angle of the ith flicker component, τ s represents the sampling period.

3. The method for estimating the on-site signal flicker parameter according to claim 1, wherein, Specifically, the estimated value of the state vector of the on-site signal at time k is calculated based on the following formula: Among them, represents the estimated value of the state vector at time k, φ represents the state transition matrix, which is a constant identity matrix, represents the estimated value of the state vector at time k - 1, Z k represents the actual measured value of the on-site signal at time k, H k represents the observation matrix, K k represents the gain matrix, K k = P k-1 (I - αP k-1 +(H k ) T R -1 H k P k ) -1 φ T (H k ) T R -1 , P k represents the covariance matrix of the estimation error at time k, P k-1 represents the covariance matrix of the estimation error at time k - 1, I represents the identity matrix, α represents the error control factor, the superscript T represents the matrix transpose, and R represents the covariance of the noise measurement during the on-site signal acquisition.

4. The method for estimating the on-site signal flicker parameter according to claim 3, characterized in that The covariance matrix P of the estimation error at time k k During the estimation process, it is updated in a recursive form, and the update expression is: P k = φP k-1 (I - αP k-1 +(H k ) T R -1 H k P k-1 ) -1 φ T 。 5. The method for estimating the on-site signal flicker parameter according to claim 3, wherein Estimation calculations are performed based on the following formula to obtain multiple envelopes at time k: Among them, represents the estimated value of the envelope corresponding to the nth harmonic component of the on-site signal at time k, represents the 2nth element in the estimated value of the state vector at time k, represents the (2n + 1)th element in the estimated value of the state vector at time k.

6. The method for estimating the flicker parameter of on-site signals according to claim 1, characterized in that, The process of inputting the multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimated values of multiple flicker components in the on-site signal specifically includes the following: Take the N envelopes at time k as the input of the adaptive linear perceptron neural network, establish the weight vectors corresponding to the N envelopes at time k, determine the method for updating the weights. After the weights of the adaptive linear perceptron neural network are updated, output the amplitude estimation value of the i-th flickering component in the on-site signal: denotes the amplitude estimation value of the i-th flickering component in the on-site signal, denotes the amplitude estimation value of the fundamental wave component in the on-site signal, denotes the (2i + 1)-th element in the weight vector, denotes the (2i + 2)-th element in the weight vector.

7. The method for estimating the flicker parameter of on-site signals according to claim 6, characterized in that, The method for updating the weight is as follows: w k+1 represents the weight vector at the (k + 1)-th moment, w k represents the weight vector at the k-th moment, e k represents the tracking error, λ represents an arbitrarily small value factor, the estimated value of the state vector at the k-th moment, the superscript T represents the matrix transpose, θ k represents the adaptive learning factor at the k-th moment, β is a constant, θ 0 represents the initial learning rate.

8. An estimation system for the flicker parameters of on-site signals, characterized in that It includes: A data acquisition module for collecting the on-site signal sequence; A model construction module for constructing the linear model of the H∞ recursive estimator for the on-site signal at time k and setting the initial iteration value; An iterative operation module for inputting the collected on-site signal sequence into the linear model of the H∞ recursive estimator for iterative operation and obtaining the estimated value of the state vector of the on-site signal at time k; An envelope calculation module for performing estimation calculations based on the estimated value of the state vector at time k to obtain multiple envelopes at time k; An estimation analysis module for inputting the multiple envelopes at time k into the trained adaptive linear perceptron neural network and outputting the amplitude estimated values of multiple flicker components in the on-site signal.

9. An electronic device, characterized in that, It includes a processor and a memory. A computer program is stored in the memory. The processor is used to execute the steps of the method according to any one of claims 1 to 7 by calling the computer program stored in the memory.

10. A computer-readable storage medium for storing a computer program for estimating flicker parameters of on-site signals, characterized in that, When the computer program runs on a computer, it executes the steps of the method according to any one of claims 1 to 7.

Images