Engine inertia measurement method based on EMD algorithm

Abstract

The invention discloses an engine inertia measurement method based on an EMD algorithm, and the method comprises the steps: S1, building a three-machine nine-node model, and applying active disturbance to any node of the model; s2, measuring a power response curve of a to-be-measured unit to obtain power deviation, measuring a frequency response curve of the to-be-measured unit, and performing signal decomposition through EMD to obtain an intrinsic mode function of the to-be-measured unit; s3, extracting a reconstruction signal in the intrinsic mode function of the unit to be tested; and S4, calculating the engine inertia of the to-be-measured unit according to the reconstructed signal. According to the invention, the influence of frequency fluctuation caused by multi-machine oscillation on an inertia measurement result can be avoided, in addition, the algorithm does not need to preset a primary function, various complex frequency signals can be processed, and the flexibility and robustness of frequency signal processing are enhanced.

Description

Technical Field

[0001] The invention relates to the technical field of industrial measurement, and in particular to an engine inertia measurement method based on an EMD algorithm. Background Art

[0002] With the large-scale integration of renewable energy and the widespread application of power electronic equipment, the traditional structure and operating characteristics of power systems have undergone significant changes. These renewable energy generation methods usually have lower system inertia than traditional synchronous generators, resulting in a decrease in the inertia level of the overall power system. As one of the key parameters for maintaining grid frequency stability, system inertia affects the system's ability to recover in the face of disturbances. Therefore, the reduction in inertia level leads to a negative impact on the frequency stability of the power system, increasing the risk of system instability. In order to ensure the safe and stable operation of the power system, the specific impact of renewable energy access on system frequency stability should be better understood through inertia measurement and analysis, and then corresponding control strategies and optimization measures should be taken. However, the multi-machine oscillation phenomenon may interfere with the measurement and evaluation of inertia because it causes abnormal fluctuations in system frequency. In order to avoid this impact, advanced measurement techniques and algorithms are needed to accurately identify and separate the interference of multi-machine oscillation on inertia measurement, so as to obtain more accurate system inertia data. This will not only help improve the frequency stability of the power system, but will also provide important technical support for the further access of renewable energy and the sustainable development of the power grid.

[0003] As the inertia of the power system decreases, the system becomes more susceptible to frequency instability, so it is critical to accurately measure and evaluate the system inertia and ensure frequency response. There are many ways to measure inertia: 1. Using the node frequency response to different disturbance levels of the power system, a method is proposed to analyze the inertial dynamics and extract features from the frequency changes of subtle disturbances, which can be used to track the inertial characteristics of the system in real time by visualization; 2. Use wavelet multi-resolution analysis technology to process the measurement signal, separate the actual signal from the interference, accurately obtain the vibration system parameters, and identify the interference source and equipment defects. However, the wavelet analysis method requires the selection of appropriate wavelet functions and decomposition levels, and different operating conditions need to be considered separately, which is not universally applicable. 3. The controlled autoregressive identification model and the time-varying forgetting factor recursive least squares algorithm are used to estimate the equivalent virtual inertia of the wind farm, thus achieving fast and accurate estimation of the equivalent virtual inertia of the wind farm; 4. Based on the dynamic response index measurement method of nonlinear least squares curve fitting, a self-verifiable system moment of inertia and damping coefficient comprehensive measurement method is established through time domain response analysis in off-grid and grid-connected states. However, this method depends on specific operating states and conditions and is not applicable to all types of power systems; 5. The regional inertia assessment technology based on the difference calculation method evaluates the regional inertia by comparing the changes in the tie line power and the dominant unit frequency after the disturbance. This technology requires less data and is suitable for online implementation, but the method has low anti-interference ability. If the measurement data of the tie line power and the dominant unit frequency are interfered by noise, it may affect the assessment results of the difference calculation method; 6. Real-time frequency partitioning and inertia parameter identification method, using dynamic time bending index and k-medoids algorithm to perform online clustering of node frequency response curves, and combining Kalman filtering to ensure real-time calculation of inertia parameters, and analyzing the impact of error sources on partitioning and identification results. This method will be affected by the selected characteristics and clustering strategies and is not suitable for all types of power grid structures or operating conditions.

[0004] It can be seen that the existing technologies all have more or less interference problems or adaptability problems. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, the engine inertia measurement method based on the EMD algorithm provided by the present invention solves the problems that the wavelet transform requires artificial selection of the wavelet basis, is limited by the Heisenberg uncertainty principle, needs to sacrifice frequency accuracy when improving time accuracy, and has poor adaptability to non-stationary and nonlinear signals.

[0006] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is: an engine inertia measurement method based on the EMD algorithm, comprising: S1. Establish a three-machine nine-node model and apply active disturbance to any node in the model; S2. For the unit to be tested, measure its power response curve to obtain the power deviation, measure its frequency response curve, and perform signal decomposition through EMD to obtain the intrinsic mode function of the unit to be tested; S3, extracting the reconstructed signal in the intrinsic mode function of the unit to be tested; S4. Calculate the engine inertia of the unit to be tested according to the reconstructed signal.

[0007] Further: S2 includes: S201, obtaining a synchronization signal of the unit to be tested and a current node signal; S202, checking whether the synchronization signal and the node signal are monotonic; S203, finding the local maximum and minimum values, and obtaining the maximum point and the minimum point; S204, interpolating the maximum value point and the minimum value point respectively to obtain an upper envelope and a lower envelope; S205: Obtaining envelope mean value according to upper envelope and lower envelope m ( t ); S206, use current signal r ( t ) and envelope mean m ( t ) to obtain a new signal h ( t ); S207, judging new signals h ( t ) whether it is monotonous; If yes, proceed to S208; If not, the new signal h ( t ) as the new current signal r ( t ), and return to S203; S208, the new signal h ( t ) as the intrinsic mode function of the current node; S209, the original signal r’ ( t ) and new signal h ( t ) to obtain the new current signal r ( t ) and then return to S203; S210, repeatedly executing S201-S210 until the intrinsic mode function of each node of the unit to be tested is obtained.

[0008] Further: in S201, the engine synchronization signal is directly acquired through the engine, and the current node signal is acquired through a phase-locked loop.

[0009] Further: In S3, the method for extracting the reconstructed signal in the intrinsic mode function of each node of the unit to be tested is: subtracting the intrinsic mode function of each node from the original signal of the disturbed engine in turn to obtain the reconstructed signal in the intrinsic mode function of each node of the unit to be tested.

[0010] Further: In S4, the engine inertia is calculated H The mathematical expression is:

[0011] in, is the engine frequency change rate, represented by the reconstructed signal, is the power change, f 0 is the rated frequency of the system, It is the rated capacity or base capacity.

[0012] The beneficial effects of the present invention are: 1. An engine inertia measurement method based on the EMD algorithm is proposed, which can identify the real dynamic behavior of the system and avoid the influence of frequency fluctuations caused by multi-machine oscillation on the inertia measurement results; 2. No need to preset basis functions, it can process various complex frequency signals, enhancing the flexibility and robustness of frequency signal processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a flow chart of the engine inertia measurement method based on the EMD algorithm.

[0014] Figure 2 It is a schematic diagram of measuring the inertia of the engine 2.

[0015] Figure 3 The frequency change curve comparison of engine 2 before and after using the EMD algorithm and the comparison chart of the maximum absolute value point of the frequency change rate are shown.

[0016] Figure 4 This is the intrinsic mode function signal and reconstructed signal diagram of engine 2 after the frequency curve is decomposed by the EMD algorithm.

[0017] Figure 5 It is a schematic diagram of measuring the inertia of the engine 3.

[0018] Figure 6 The frequency change curves of engine 3 before and after using the EMD algorithm are compared, as well as the absolute maximum point comparison of the frequency change rate.

[0019] Figure 7 This is the intrinsic mode function signal and reconstructed signal diagram of engine 3 after the frequency curve is decomposed by the EMD algorithm. DETAILED DESCRIPTION

[0020] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0021] like Figure 1 As shown, in one embodiment of the present invention, an engine inertia measurement method based on an EMD algorithm is provided, comprising: S1. Establish a three-machine nine-node model and apply active disturbance to any node in the model; The method of applying active disturbance is: applying active disturbance to the node after the system enters a stable state, that is, connecting a certain size of active load; S2. For the unit to be tested, measure its power response curve to obtain the power deviation, measure its frequency response curve, and perform signal decomposition through EMD to obtain the intrinsic mode function of the unit to be tested; S3, extracting the reconstructed signal in the intrinsic mode function of the unit to be tested; S4. Calculate the engine inertia of the unit to be tested according to the reconstructed signal.

[0022] S2 includes: S201, obtaining the synchronization signal of the unit to be tested and the current node signal; S202, checking whether the synchronization signal and the node signal are monotonic; S203, finding the local maximum and minimum values, and obtaining the maximum point and the minimum point; S204, interpolating the maximum value point and the minimum value point respectively to obtain an upper envelope and a lower envelope; S205: Obtaining envelope mean value according to upper envelope and lower envelope m ( t ); S206, use current signal r ( t ) and envelope mean m ( t ) to obtain a new signal h ( t ); S207, judging new signals h ( t ) whether it is monotonous; If yes, proceed to S208; If not, the new signal h ( t ) as the new current signal r ( t ), and return to S203; S208, the new signal h ( t ) as the intrinsic mode function of the current node; S209, the original signal r’ ( t ) and new signal h ( t ) to obtain the new current signal r ( t ) and then return to S203; Among them, the original signalr’ ( t ) means the frequency signal of the unit under test after active disturbance is applied; S210, repeatedly executing S201-S210 until the intrinsic mode function of each node of the unit to be tested is obtained.

[0023] In S201, the engine synchronization signal is directly obtained through the engine, and the current node signal is obtained through a phase-locked loop.

[0024] Furthermore, in S3, the method for extracting the reconstructed signal in the intrinsic mode function of each node of the unit to be tested is: subtracting the intrinsic mode function of each node from the original signal of the disturbed engine in turn to obtain the reconstructed signal in the intrinsic mode function of each node of the unit to be tested.

[0025] In S4, calculate the engine inertia H The mathematical expression is:

[0026] in, is the engine frequency change rate, represented by the reconstructed signal, is the power change, f 0 is the rated frequency of the system, It is the rated capacity or base capacity.

[0027] Calculate engine inertia H Another expression is:

[0028] in, J is the moment of inertia, is the rated mechanical angular velocity of the unit; like Figure 2 , Figure 3 and Figure 4 As shown, in this embodiment, the B8 node is applied The disturbance, Figure 2 middle P Li Indicates load, B i represents the busbar, G i Indicates generator T i Represents a transformer, i For counting identification, the No. 2 engine group was studied, and the corresponding frequency change rate curve for the first two seconds (represented by the blue line) and the curve after decomposition by the EMD algorithm (represented by the red line) were obtained; using the original curve for calculation, we can get The maximum absolute value is 15.236179, corresponding to t= 0.000050, and the inertia is calculated H The value is 0.7129; using the decomposed curve to calculate, we can get The maximum absolute value is 3.252742, which is obtained when t=0.349950. The inertia is calculated H The value of 3.3394 is consistent with the set inertia H The value 3.33 is closer.

[0029] like Figure 5 , Figure 6 and Figure 7 As shown, in another embodiment of the present invention, the B3 node is applied The disturbance of the No. 3 engine unit is studied and the original curve is used for calculation, and it can be obtained The maximum absolute value is 0.47Hz / s, so Value 0.28s; using the decomposed curve to calculate, we can get The maximum absolute value is 0.056Hz / s, so The value is 2.35s, which is consistent with the setting The value of 2.35s is consistent.

[0030] The analysis shows that the original curve has signal distortion. The nonlinear factors introduced by multi-machine oscillation make the signal show irregular fluctuations in the time domain. These fluctuations are reflected as unstable frequency components in the frequency domain. When the original curve is subjected to frequency analysis, it is observed that the maximum value of the frequency change rate is abnormally prominent at certain points, which is mainly due to the existence of noise fluctuation points. These abnormal points do not represent the real frequency changes of the signal, but are illusions caused by signal distortion. In order to improve the accuracy of frequency analysis, the EMD algorithm is used to denoise the original curve to reduce the influence of noise on the frequency characteristics.

[0031] By calculating the corresponding engine inertia H , and compared with the set parameters, it is proved that the EMD algorithm can get accurate values, otherwise the interference of multi-machine oscillation will seriously affect The accuracy of the value affects the inertia H In addition, the EMD algorithm can also be used to quickly decompose the abnormal oscillation part of the signal, that is, Figure 4 The IMF1 signal in the system helps to identify and locate the source of abnormalities in the system, providing important direction for fault diagnosis and system maintenance.

[0032] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. The engine inertia measurement method based on the EMD algorithm is characterized by: include: S1. Establish a three-machine nine-node model and apply active disturbance to any node in the model; S2. For the unit to be tested, measure its power response curve to obtain the power deviation, measure its frequency response curve, and perform signal decomposition through EMD to obtain the intrinsic mode function of the unit to be tested; S3, extracting the reconstructed signal in the intrinsic mode function of the unit to be tested; S4. Calculate the engine inertia of the unit to be tested according to the reconstructed signal.

2. The engine inertia measurement method based on the EMD algorithm according to claim 1, characterized in that: The S2 includes: S201, obtaining the synchronization signal of the unit to be tested and the current node signal; S202, checking whether the synchronization signal and the node signal are monotonic; S203, finding the local maximum and minimum values, and obtaining the maximum point and the minimum point; S204, interpolating the maximum value point and the minimum value point respectively to obtain an upper envelope and a lower envelope; S205: Obtaining envelope mean value according to upper envelope and lower envelope m ( ts ); S206, use current signal r ( t ) and envelope mean m ( t ) to obtain a new signal h ( t ); S207, judging new signals h ( t ) whether it is monotonous; If yes, proceed to S208; If not, the new signal h ( t ) as the new current signal r ( t ), and returns to S203; S208, the new signal h ( t ) as the intrinsic mode function of the current node; S209, the original signal r’ ( t ) and new signal h ( t ) to obtain the new current signal r ( t ) and then return to S203; S210, repeatedly executing S201-S210 until the intrinsic mode function of each node of the unit to be tested is obtained.

3. The engine inertia measurement method based on the EMD algorithm according to claim 2, characterized in that: In S201, the synchronization signal of the unit to be tested is directly obtained through the engine, and the current node signal is obtained through a phase-locked loop.

4. The engine inertia measurement method based on the EMD algorithm according to claim 1, characterized in that: In S3, the method for extracting the reconstructed signal in the intrinsic mode function of each node of the unit under test is: subtracting the intrinsic mode function of each node from the original signal of the unit under test in turn to obtain the reconstructed signal in the intrinsic mode function of each node of the unit under test.

5. The engine inertia measurement method based on the EMD algorithm according to claim 1, characterized in that: In S4, the engine inertia is calculated H The mathematical expression is: in, is the engine frequency change rate, represented by the reconstructed signal, is the power change, f 0 is the rated frequency of the system, It is the rated capacity or base capacity.

Images