Method for evaluating inertia of regional power grid containing high-proportion new energy based on data driving

Abstract

A data-driven regional power grid inertia evaluation method containing high-proportion new energy comprises the following steps: analyzing multi-modal distribution of active power and frequency of different nodes by using a recursive typicality data analysis method, calculating inertia center frequency, and calculating inertia center frequency; identifying a representative node of the regional power grid with the frequency response closest to the inertia center frequency in the regional power grid; obtaining norm indexes of the active power and the frequency of the representative inter-node bus of the regional power grid according to the active power and the frequency of the representative node of the regional power grid; the regional power grid inertia is evaluated by using an autoregressive moving average input model in combination with norm indexes of active power and frequency of regional representative inter-node buses, and an important regulation and control basis is provided for stable operation of the regional power grid. The method has the advantages that the TDA model and the ARMAX model of the method can realize real-time and dynamic evaluation of the regional power grid inertia without depending on the traditional synchronous generator average inertia hypothesis and power grid detailed topological information.

Description

Technical Field

[0001] The present invention relates to a data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy. Background Art

[0002] With the transformation of the global energy structure and the improvement of environmental protection awareness, new energy generation, such as solar energy and wind energy, is being connected to the power grid at an unprecedented rate. This trend not only promotes the diversification of the energy structure, but also provides strong support for achieving sustainable development goals. However, the large-scale connection of new energy to the grid has also brought new challenges and problems to the operation and management of the power grid.

[0003] Renewable energy generation usually relies on power electronics to connect to the grid, which is significantly different from traditional synchronous generators. Traditional synchronous generators contain rotating mass, which can provide the necessary inertia support when the grid frequency changes, contributing to the stable operation of the grid. However, power electronics for renewable energy generation usually do not have rotating mass, so they cannot provide effective inertia support like synchronous generators.

[0004] In addition, power electronic devices have the characteristics of fast response and can quickly adjust the output power according to the needs of the power grid. While this fast response mechanism improves the flexibility and efficiency of the power grid, it may also cause the power grid to lack sufficient inertia support when it is disturbed. When the power grid is disturbed, if there is not enough inertia to absorb and slow down the frequency change, it may cause the instability of the power grid frequency and even cause the power grid to collapse. This phenomenon is called grid inertia displacement, which is a new problem faced by the power grid after the large-scale grid connection of new energy.

[0005] To address this problem, an accurate assessment of the inertia of the power grid is required. However, traditional methods have many deficiencies when it comes to inertia assessment. On the one hand, analytical calculation methods based on knowledge of generator inertia and frequency estimation require topological information of the power grid, and when renewable energy generation is added on a large scale, the accuracy of these methods will be affected, resulting in inaccurate frequency calculations. On the other hand, traditional methods also rely on accurate system models and power grid topology information, which may lead to inaccurate inertia assessment when the power grid structure changes frequently or the information is incomplete.

[0006] Therefore, it is particularly important to develop a new method that can adapt to the large-scale grid connection of renewable energy, does not require precise system models and grid topology information, and can accurately evaluate grid inertia. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy. The method adopts typicality data analysis (TDA) and autoregressive moving average exogenous input (ARMAX) model, and can realize real-time and dynamic evaluation of the inertia of the regional power grid without relying on the traditional assumption of the average inertia of synchronous generators and detailed topology information of the power grid.

[0008] The technical solution of the present invention is:

[0009] A data-driven method for evaluating the inertia of a regional power grid with a high proportion of renewable energy, comprising:

[0010] S1: Collect active power and frequency measurement data from the regional power grid containing renewable energy;

[0011] S2: Using the recursive typicality data analysis (TDAp) method, the multimodal distribution of active power and frequency of different nodes is analyzed, the inertia center frequency is calculated, and the representative nodes of the regional power grid whose frequency response is closest to the inertia center frequency are identified;

[0012] S3: Obtain the norm index of the active power and frequency of the representative internode bus of the regional power grid according to the active power and frequency of the representative node of the regional power grid;

[0013] S4: The autoregressive moving average input (ARMAX) model is used to evaluate the inertia of the regional power grid, combining the norm indicators of the active power and frequency of the regional representative internode buses, providing an important control basis for the stable operation of the regional power grid.

[0014] Furthermore, in step S2, the inertia center frequency f coi The calculation formula of (t) is:

[0015]

[0016] In the formula,

[0017]

[0018] in,

[0019] N g is the number of thermal generators and synchronous condensers in the area;

[0020] N m is the number of transmission buses to which all renewable energy generators in the region are connected;

[0021] N μ is the number of all renewable energy generators in the region;

[0022] f n (t) and inertia Hn are the frequency and inertia of the thermal generator and synchronous condenser respectively;

[0023] f m (t) the frequency of the transmission bus to which a considerable number of renewable energy generators are connected;

[0024] H m The inertia on the transmission bus connected to the new energy generator;

[0025] h i and ω i are the inertia and speed of each new energy generator respectively;

[0026] ω eq For all new energy generators on the transmission bus (N μ ) equivalent speed.

[0027] Furthermore, in step S3, the frequency deviation and active power deviation of the bus between representative nodes of the regional power grid are calculated by using the frequency and active power of the representative nodes respectively, and the Euclidean norm of the frequency and active power of the bus between representative nodes is obtained.

[0028] Further preferably, the Euclidean norm of the frequency and active power of the inter-node bus is in the form of a distribution space, and the norm index α(k, j) of the active power and frequency of the representative inter-node bus of the regional power grid is expressed as:

[0029] α(k,j)={[β(k,1);π(k,1)],[β(k,2);π(k,2)],…,[β(k,j);π(k,j)]};

[0030] in,

[0031] β(k,j) is the frequency deviation of the bus between the representative nodes k and j of the regional power grid;

[0032] π(k,j) is the active power deviation between the buses of the representative nodes k and j in the regional power grid

[0033] Furthermore, the autoregressive moving average input model is a high-order autoregressive moving average input model.

[0034] Further preferably, when evaluating the inertia of the regional power grid, the poles of the transfer function of the high-order autoregressive moving average input model are obtained, the tenth-order autoregressive moving average input (ARMAX) model of the poles is reduced to a first-order autoregressive moving average input (ARMAX) model using a MATLAB function, and the first-order transfer function G(s) is obtained, and the inertia of the regional power grid is inferred based on the first-order transfer function G(s).

[0035] The beneficial effects of the present invention are:

[0036] This method adopts typical data analysis (TDA) and autoregressive moving average exogenous input (ARMAX) model, without relying on traditional synchronous generator average inertia assumptions and detailed grid topology information, to achieve real-time and dynamic evaluation of regional grid inertia. It has strong adaptability and high accuracy, and can achieve accurate evaluation of regional grid inertia in renewable energy power systems, providing strong support for the stable operation and optimized scheduling of power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0038] Example 1

[0039] The data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy sources of the present invention includes steps S1-S4, which are specifically as follows:

[0040] S1: Collect active power and frequency measurement data from the regional power grid containing renewable energy;

[0041] S2: Using the recursive typicality data analysis (TDAp) method, the multimodal distribution of active power and frequency of different nodes is analyzed, the inertia center frequency is calculated, and the representative nodes of the regional power grid whose frequency response is closest to the inertia center frequency are identified;

[0042] Among them, the inertia center frequency f coi (t) is:

[0043]

[0044] In the formula,

[0045]

[0046] in,

[0047] N g is the number of thermal generators and synchronous condensers in the area;

[0048] N m is the number of transmission buses to which all renewable energy generators in the region are connected;

[0049] N μ is the number of all renewable energy generators in the region;

[0050] f n (t) and inertia H n are the frequency and inertia of the thermal generator and synchronous condenser respectively;

[0051] f m(t) the frequency of the transmission bus to which a considerable number of renewable energy generators are connected;

[0052] H m The inertia on the transmission bus connected to the new energy generator;

[0053] h i and ω i are the inertia and speed of each new energy generator respectively;

[0054] ω eq For all new energy generators on the transmission bus (N μ ) equivalent speed;

[0055] By using the typical data analysis (TDA) method, the frequency of a representative node of the regional power grid can be approximated; the frequency of this node is the inertia center frequency:

[0056] f coi (t)≈f(t)(3)

[0057] in,

[0058] f(t) is the frequency of the representative node of the regional power grid;

[0059] The recursive typicality data analysis (TDAp) method is used to obtain the frequencies of representative nodes of a regional power grid with a limited number of nodes, making data measurement more intuitive and improving the recognition efficiency of the ARMAX model.

[0060] S3: Obtaining the norm index of the active power and frequency of the representative internode bus of the regional power grid according to the active power of the representative node and the frequency of the node;

[0061] (1) Obtaining the frequency deviation of representative nodes

[0062] Frequency deviation Δf of representative nodes k and j in regional power grid k (t) and Δf j (t) are:

[0063]

[0064] in,

[0065] f k (t) is the node frequency of node k at time t from the start of the disturbance;

[0066] f k (t0) is the node frequency of node k at time t0;

[0067] f j (t) is the node frequency of node j at time t from the start of the disturbance;

[0068] f j (t0) is the node frequency of node j at time t0;

[0069] (2) Obtaining the frequency deviation of the bus between representative nodes

[0070] The frequency deviation β(k,j) of the bus between the representative nodes k and j of the regional power grid is:

[0071]

[0072] in,

[0073] Δf k (t) is the frequency deviation of node k from the disturbance moment;

[0074] Δf j (t) is the frequency deviation of node j from the disturbance moment;

[0075] t f is the disturbance moment;

[0076] (3) Obtaining active power deviation of representative nodes

[0077] Active power deviation Δpe of representative nodes k and j in the regional power grid k (t) and Δpe j (t) are:

[0078]

[0079] in,

[0080] pe k (t0) is the active power of node k at time t0;

[0081] pe k (t) is the active power of node k at time t from the start of the disturbance;

[0082] pe j (t0) is the active power of node j at time t0;

[0083] pe j (t) is the active power of node j at time t from the start of the disturbance;

[0084] (4) Obtaining active power deviation

[0085] The active power deviation π(k, j) of the busbar between the representative nodes k and j of the regional power grid is:

[0086]

[0087] in,

[0088] △Pek (t) is the active power deviation of node k from the start of the disturbance;

[0089] Δpe j (t) is the active power deviation of node j from the start of the disturbance;

[0090] (5) The norm index of active power and frequency of representative internode buses of regional power grid in the form of distribution space The norm index α(k, j) of active power and frequency of representative internode buses of regional power grid in the form of distribution space is:

[0091] α(k,j)={[β(k,1);π(k,1)],[β(k,2);π(k,2)],…,[β(k,j);π(k,j)]} (10)

[0092] S4: Using the autoregressive moving average input (ARMAX) model, combined with the norm indicators of active power and frequency of the regional representative internode bus, the regional power grid inertia is evaluated, providing an important control basis for the stable operation of the regional power grid;

[0093] In order to evaluate the inertia of the regional power grid, the norm indicators of active power and frequency of the representative internode bus of the regional power grid are combined as model inputs, and the tenth-order autoregressive moving average input (ARMAX) model is used. The high-order (tenth-order) autoregressive moving average input model is used to improve the evaluation accuracy of the dynamic behavior of the regional power grid during internode disturbances.

[0094] Through the error (NPSE), determine the poles of the transfer function;

[0095]

[0096] in,

[0097] f coi (t) is the center frequency of inertia;

[0098] f pb (t) is the frequency response of the node;

[0099] The MATLAB function is used to reduce the tenth-order autoregressive moving average input (ARMAX) model of the pole to a first-order autoregressive moving average input (ARMAX) model, where the first-order transfer function G(s) is obtained as follows:

[0100] The frequency response of the node is approximately the center frequency of inertia. for:

[0101]

[0102] is the total deviation of the active power of the node;

[0103] Frequency response of the node;

[0104] H Req is the estimated inertia of the regional power grid;

[0105] By performing Laplace transform on formula (12), the time domain nodes in formula (12) are converted into inertia responses of frequency domain nodes, and the first-order transfer function G(s) is obtained:

[0106]

[0107] in,

[0108] Δf pb (s) is Δf k (t) Frequency deviation of the Laplace transform;

[0109] ΔPe B (s) is Δpe k (t) Laplace transformed active power deviation;

[0110] D is the differential operator, s is the Laplace transform coefficient;

[0111] According to the first-order transfer function G(s), the regional power grid inertia is inferred as:

[0112]

[0113] In summary, the present invention collects data from a regional power grid containing new energy, analyzes data, evaluates the dynamic behavior of the power grid, and estimates the system inertia. It has strong adaptability and high accuracy, and provides important technical support for the stable operation of the power grid.

[0114] The above are only specific embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy, characterized by: include: S1: Collect active power and frequency measurement data from the regional power grid containing renewable energy; S2: Using the recursive typicality data analysis method, analyze the multimodal distribution of active power and frequency of different nodes, calculate the inertia center frequency, and identify the representative nodes of the regional power grid whose frequency response is closest to the inertia center frequency; S3: Obtain the norm index of the active power and frequency of the representative internode bus of the regional power grid according to the active power and frequency of the representative node of the regional power grid; S4: The autoregressive moving average input model is used in combination with the norm indicators of active power and frequency of the regional representative internode buses to evaluate the inertia of the regional power grid, providing an important control basis for the stable operation of the regional power grid.

2. The data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy sources according to claim 1 is characterized in that: In step S2, the inertia center frequency f coi The calculation formula of (t) is: In the formula, in, N g is the number of thermal generators and synchronous condensers in the area; N m is the number of transmission buses to which all renewable energy generators in the region are connected; N μ is the number of all renewable energy generators in the region; f n (t) and inertia H n are the frequency and inertia of the thermal generator and synchronous condenser respectively; f m (t) the frequency of the transmission bus to which a considerable number of renewable energy generators are connected; H m The inertia on the transmission bus connected to the new energy generator; h i and ω i are the inertia and speed of each new energy generator respectively; ω eq For all new energy generators on the transmission bus (N μ ) equivalent speed.

3. The data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy sources according to claim 1 is characterized in that: In step S3, the frequency deviation and active power deviation of the bus between representative nodes of the regional power grid are calculated by using the frequency and active power of the representative nodes respectively, and the Euclidean norm of the frequency and active power of the bus between representative nodes is obtained.

4. The data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy sources according to claim 3 is characterized in that: The Euclidean norm of the frequency and active power of the inter-node bus is in the form of distribution space, and the norm index α(k, j) of the active power and frequency of the representative inter-node bus of the regional power grid is expressed as: α(k,j)={[β(k,1);π(k,1)],[β(k,2);π(k,2)],…,[β(k,j);π(k,j)]}; in, β(k,j) is the frequency deviation of the bus between the representative nodes k and j of the regional power grid; π(k,j) is the active power deviation between the buses of the representative nodes k and j in the regional power grid.

5. The data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy sources according to claim 1 is characterized in that: The autoregressive moving average input model is a high-order autoregressive moving average input model.

6. The data-driven method for evaluating the inertia of a regional power grid containing a high proportion of new energy sources according to claim 5 is characterized by: When evaluating the inertia of the regional power grid, the poles of the transfer function of the high-order autoregressive moving average input model are obtained, and the MATLAB function is used to reduce the tenth-order autoregressive moving average input model of the poles to a first-order autoregressive moving average input model, and the first-order transfer function G(s) is obtained. The inertia of the regional power grid is inferred based on the first-order transfer function G(s).

Images