A power system source-network-load-storage collaborative intelligent planning system

Abstract

The application relates to the technical field of power systems, and discloses a power system source-network-load-storage collaborative intelligent planning system, which obtains the power grid load value of a target power grid at fixed time intervals; carries out discrete wavelet transform processing on the time-series power grid load value to obtain M time-scale wavelet coefficients, and constructs a corresponding fractal elastic demand kernel in combination with a preset fractal order; obtains the energy storage value of each energy storage device in the nth physical node and the phase angle of each energy storage device at fixed time intervals, and determines the inertia vector phase field of the nth physical node in combination with a preset angular frequency; constructs a synchronous offset matrix based on the inertia vector phase field of each physical node and the fractal elastic demand kernel of each time scale; and constructs a target function in combination with the synchronous offset matrix based on the energy storage unit cost and the response unit cost, so as to determine the energy storage configuration and demand response strategy of each physical node.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and more specifically, to a power system source-grid-load-storage coordinated intelligent planning system. Background Technology

[0002] In power systems with a high proportion of renewable energy integration, the load side and the inertia support side exhibit two characteristics, namely fractal elastic demand and multi-source inertia vector phase field.

[0003] Fractal elastic demand: The load response to price signals or regulatory commands exhibits significant self-similar fractal characteristics, with its power adjustment gradient following a non-integer power law on timescales from seconds to minutes (fractal order). This characteristic leads to multi-scale randomness in load fluctuations, causing the system inertia demand to exhibit a non-uniform distribution in the time domain.

[0004] Multi-source inertia vector phase field: The spatial field distribution is formed by the superposition of equivalent inertia vectors from multiple inertia sources such as distributed inverters and energy storage devices. The energy state and control phase of each inertia source constitute the vector basis. The superposition result exhibits a significant anisotropic phase coupling effect in weak links of the power grid topology (such as between nodes of high-impedance tie lines). The phase difference of the inertia vector will induce a dynamic amplification phenomenon similar to field oscillation, which directly affects the spatial distribution characteristics of the system frequency.

[0005] The coupling effect between the aforementioned fractal elasticity requirement and the multi-source inertia vector phase field leads to a mismatch between inertia and elastic response, specifically manifested as follows:

[0006] Fractal fluctuations on the load side create a multi-scale stochastic power gap;

[0007] The inertial field is affected by topological impedance and equipment control delay, and the phase adjustment speed cannot match the load fluctuation frequency, resulting in a double lag in inertial support in the spatiotemporal dimensions.

[0008] Traditional power system planning methods only consider the linear response of the load and the centralized inertia support mode, without establishing a correlation model between fractal elastic loads and multi-source phase angle fields. This makes it impossible to achieve multi-scale collaborative optimization. Specifically, the shortcomings are as follows: in scenarios with a high proportion of power electronic equipment connected, the system frequency control accuracy is insufficient, weak nodes in the topology are prone to frequency deviation exceeding the limit, and the uncertainty of inertia leads to redundancy in reserve capacity configuration, which exacerbates the cost of new energy consumption. Summary of the Invention

[0009] This invention provides a power system source-grid-load-storage coordinated intelligent planning system, which solves the technical problems mentioned in the background art.

[0010] This invention provides a power system source-grid-load-storage coordinated intelligent planning system, comprising:

[0011] The first acquisition module is used to acquire the grid load value of the target power grid at fixed time intervals;

[0012] The first processing module is used to perform discrete wavelet transform processing on the time-series grid load values ​​to obtain wavelet coefficients of M time scales, and to construct the corresponding fractal elastic demand kernel by combining the preset fractal order.

[0013] The second acquisition module is used to acquire the energy storage value of each energy storage device in the nth physical node at fixed time intervals, as well as the phase angle of each energy storage device, and to determine the inertia vector phase field of the nth physical node in combination with the preset angular frequency; 1≤n≤N, where n is a positive integer;

[0014] The second processing module is used to construct a synchronization offset matrix based on the inertia vector phase field of each physical node and the fractal elastic demand kernel of each time scale.

[0015] The intelligent planning module is used to construct an objective function based on the unit cost of energy storage and the unit cost of response, combined with the synchronization offset matrix, to determine the energy storage configuration and demand response strategy for each physical node.

[0016] Furthermore, the time-series grid load values ​​are subjected to discrete wavelet transform processing to obtain wavelet coefficients at M time scales, including:

[0017] A load sequence is constructed based on the time-series grid load values, and each grid load value in the load sequence is normalized by maximum and minimum values ​​to form a net load sequence.

[0018] Set M time scales , satisfy: ,and ;in, Represents a fixed time interval, for each time scale The wavelet coefficients are obtained by calculating the discrete wavelet transform. , It is a positive integer.

[0019] Furthermore, a corresponding fractal elasticity requirement kernel is constructed by combining a preset fractal order, including:

[0020] For each time scale The absolute values ​​of the corresponding wavelet coefficients are calculated. ;

[0021] Based on fractal order and time scale Calculate the corresponding scale weights , ;

[0022] For each time scale , absolute value With scale weight According to fractal order Coupling forms fractal elastic fundamental quantities;

[0023] The fractal elasticity basic quantities are globally normalized to output the fractal elasticity demand kernel.

[0024] Furthermore, the energy storage value of each energy storage device within the nth physical node and the phase angle of each energy storage device are acquired at fixed time intervals. Combined with a preset angular frequency, the inertia vector phase field of the nth physical node is determined, including:

[0025] Based on the energy storage value and phase angle of the energy storage device, the inertia vector component of the corresponding energy storage device is calculated in combination with the preset angular frequency.

[0026] The inertia vector phase field of the nth physical node is obtained by summing the inertia vector components of all energy storage devices within the nth physical node.

[0027] Furthermore, based on the inertia vector phase field of each physical node and the fractal elasticity requirement kernel of each time scale, a synchronization offset matrix is ​​constructed, including:

[0028] Regarding the first At each time point, determine the inertia magnitude of the nth physical node, and the inertia magnitude at each time scale. The fractal elastic demand kernel; where the inertia amplitude of the nth physical node is the magnitude of the corresponding inertia vector phase field;

[0029] The inertia magnitude of the nth physical node is expanded into a first vector of dimension M through M copies;

[0030] Based on M time scales The fractal elastic demand kernel is used to construct a second vector of dimension M;

[0031] Calculate the element-wise ratio of the first vector to the second vector to obtain the mismatch rate;

[0032] The mismatch rate is calculated using the natural logarithm and its absolute value is taken to form the synchronization offset value.

[0033] A synchronization offset matrix is ​​constructed based on the synchronization offset values ​​of the timing of N physical nodes.

[0034] Furthermore, based on the unit cost of energy storage and the unit cost of response, an objective function is constructed by combining the synchronization offset matrix, including:

[0035] Load and generate the initial energy storage capacity planning quantity and the initial response depth planning quantity for the nth physical node;

[0036] For the nth physical node, the energy storage cost is calculated based on the initial planned energy storage capacity and the unit cost of energy storage.

[0037] For the nth physical node, the response cost is calculated based on the initial response depth planning quantity and the response unit cost;

[0038] The synchronization offset matrix is ​​reduced in dimensionality according to the physical node dimension for each time scale. And every moment Sum the offsets to obtain the offset aggregate value;

[0039] The energy storage cost and response cost of each physical node are combined to form the total economic cost item; based on the preset penalty factor... The total penalty cost is calculated by combining the offset aggregate value;

[0040] The objective function is obtained by combining the total economic cost term and the total penalty cost term.

[0041] Furthermore, the synchronization offset matrix is ​​reduced in dimensionality according to the physical node dimension for each time scale. And every moment Sum the offsets to obtain the offset aggregate value, including:

[0042] The synchronization offset matrix is ​​merged according to the dimension of the physical nodes to obtain the sum of the offsets at the time scale and at each moment; where the dimension of the physical nodes is N.

[0043] The sum of the time scale and the offset at each moment is normalized according to the number of physical nodes to obtain the average offset.

[0044] Compare the average offset with a preset offset threshold to generate an offset aggregate value, which can be 1 or 0.

[0045] Furthermore, the energy storage configuration and demand response strategy for each physical node are determined, including:

[0046] Determine the first and second value boundaries corresponding to the planned energy storage capacity and response depth for each physical node, respectively.

[0047] Set a stopping threshold for iterating the objective function value;

[0048] If the initial response depth planning quantity and the objective function value obtained from the initial response depth planning quantity for N physical nodes are less than or equal to the iteration stopping threshold, then the initial response depth planning quantity and the initial response depth planning quantity for N physical nodes are output as the energy storage configuration and demand response strategy for the corresponding physical nodes, respectively.

[0049] Otherwise, update the initial response depth planning quantity and the initial response depth planning quantity within the first value boundary and the second value boundary until the objective function value is less than or equal to the iteration stopping threshold, and output the corresponding updated response depth planning quantity and the updated response depth planning quantity as the energy storage configuration and demand response strategy of the corresponding physical node, respectively.

[0050] The beneficial effects of this invention are as follows: by constructing a power system source-grid-load-storage collaborative intelligent planning mechanism with fractal elastic demand kernel and inertia vector phase field coupling as the core, and by quantifying the spatiotemporal mismatch between load-side multi-scale response characteristics and source-side inertia support through synchronous offset matrix, and embedding this spatiotemporal mismatch into the joint optimization objective function of economic cost and response strategy, it is possible to achieve the economically optimal allocation of energy storage configuration and demand response at each physical node while ensuring system frequency stability, thereby significantly improving the renewable energy absorption capacity and system dynamic adaptability. Attached Figure Description

[0051] Figure 1 This is a system module diagram of the present invention. Detailed Implementation

[0052] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0053] like Figure 1 As shown, a power system source-grid-load-storage coordinated intelligent planning system includes:

[0054] The first acquisition module is used to acquire the grid load value of the target power grid at fixed time intervals;

[0055] The first processing module is used to perform discrete wavelet transform processing on the time-series grid load values ​​to obtain wavelet coefficients of M time scales, and to construct the corresponding fractal elastic demand kernel by combining the preset fractal order.

[0056] The second acquisition module is used to acquire the energy storage value of each energy storage device in the nth physical node at fixed time intervals, as well as the phase angle of each energy storage device, and to determine the inertia vector phase field of the nth physical node in combination with the preset angular frequency; 1≤n≤N, where n is a positive integer;

[0057] The second processing module is used to construct a synchronization offset matrix based on the inertia vector phase field of each physical node and the fractal elastic demand kernel of each time scale.

[0058] The intelligent planning module is used to construct an objective function based on the unit cost of energy storage and the unit cost of response, combined with the synchronization offset matrix, to determine the energy storage configuration and demand response strategy for each physical node.

[0059] In one embodiment of the present invention, the time-series grid load values ​​are subjected to discrete wavelet transform processing to obtain wavelet coefficients at M time scales, including:

[0060] A load sequence is constructed based on the time-series grid load values, and each grid load value in the load sequence is normalized by maximum and minimum values ​​to form a net load sequence.

[0061] Set M time scales , satisfy: ,and ;in, Represents a fixed time interval, for each time scale The wavelet coefficients are obtained by calculating the discrete wavelet transform. , It is a positive integer.

[0062] In detail, construct the load sequence ,in, , This indicates the total number of time points to establish a unified time-scale benchmark and ensure time-domain consistency in subsequent transformations.

[0063] For load sequences Perform Min-Max normalization to map the load values ​​to Interval.

[0064] Preferably, the Daubechies-4 wavelet basis function is used to denoise the net load sequence to remove high-frequency noise interference.

[0065] By setting M time scales , satisfy: ,and To form a logarithmic time scale, covering typical power system dynamics from the second to the hour.

[0066] It should be noted that the discrete wavelet transform is based on the Daubechies-4 mother wavelet to avoid the accumulation of truncation errors caused by different basis functions.

[0067] In one embodiment of the present invention, a corresponding fractal elasticity requirement kernel is constructed by combining a preset fractal order, including:

[0068] For each time scale The absolute values ​​of the corresponding wavelet coefficients are calculated. ;

[0069] It should be noted that the same set of time scales is retrieved from the discrete wavelet module. and their corresponding wavelet coefficients ; Calculate the absolute values ​​of wavelet coefficients at each scale. This converts the complex results of wavelet transform into positive quantitative values, providing a foundation for fractal modeling.

[0070] Based on fractal order and time scale Calculate the corresponding scale weights , ;

[0071] in, This is used to assign priority weights to load fluctuations at different time scales, quantifying the time scale into elastic weights to reflect fractal self-similarity characteristics. For example, short-scale (such as second-level peaks) have high weights, while long-scale (such as hourly trends) have low weights.

[0072] For each time scale , absolute value With scale weight According to fractal order Coupling forms fractal elastic fundamental quantities, as follows:

[0073]

[0074] in, Representing fractal elastic fundamental quantities to quantify time scales Under these conditions, load fluctuations can be addressed through demand response or the elastic potential of energy storage regulation. The larger the fractal elasticity base value, the more prominent the regulation demand.

[0075] The fractal elasticity fundamentals are globally normalized to output the fractal elasticity demand kernel, as follows:

[0076]

[0077] in, Let m represent the m-th time scale, where 1 ≤ m ≤ M, and m is a positive integer.

[0078] In one embodiment of the present invention, the energy storage value of each energy storage device in the nth physical node and the phase angle of each energy storage device are acquired at fixed time intervals, and the inertia vector phase field of the nth physical node is determined by combining the phase angle with a preset angular frequency, including:

[0079] Based on the energy storage value and phase angle of the energy storage device, and combined with the preset angular frequency, the corresponding inertia vector components of the energy storage device are calculated as follows:

[0080]

[0081] in, Represents the inertia vector components. This represents the energy storage value of the i-th energy storage device at time t. Indicates the preset angular frequency. This represents the phase angle of the i-th energy storage device at time t.

[0082] To assess the dynamic support that each energy storage device can provide when the system frequency changes, it is necessary to map their current energy state (energy storage value) and electrical phase (phase angle) into inertia, and inertia vector components to characterize their contribution to the overall inertia of the physical nodes.

[0083] The inertia vector phase field of the nth physical node is obtained by summing the inertia vector components of all energy storage devices within the nth physical node, as follows:

[0084]

[0085] in, This represents the phase field of the inertia vector of the nth physical node at time t. This represents the number of energy storage devices at the nth physical node.

[0086] It should be noted that the inertia contribution of a single energy storage device cannot fully reflect the dynamic characteristics of a physical node; they need to be accumulated to represent the collective response of the entire physical node to frequency fluctuations. By summing, the overall inertia vector phase field of the physical node is obtained, which is the equivalent inertia vector of the physical node in the system.

[0087] In one embodiment of the present invention, a synchronization offset matrix is ​​constructed based on the inertia vector phase field of each physical node and the fractal elastic demand kernel of each time scale, including:

[0088] Regarding the first At each time point, determine the inertia magnitude of the nth physical node, and the inertia magnitude at each time scale. The fractal elastic demand kernel; where the inertia amplitude of the nth physical node is the magnitude of the corresponding inertia vector phase field;

[0089] Phase field of the inertia vector of the nth physical node at time t By performing modulo operation, the inertia magnitude of the nth physical node at time t is obtained. .

[0090] The inertia magnitude of the nth physical node is expanded into a first vector of dimension M through M copies;

[0091] The inertia amplitude of a physical node is a single scalar and needs to be compared element-by-element with the fractal elastic demand kernel over the same length. Therefore, Amplitude M times, forming , as the first vector.

[0092] Based on M time scales The fractal elastic demand kernel is used to construct a second vector of dimension M;

[0093] Calculate the element-wise ratio of the first vector to the second vector to obtain the mismatch rate, as follows:

[0094]

[0095] in, Indicates the nth physical node in time scale The ratio of the magnitude of inertia to the elasticity of load demand.

[0096] By directly comparing supply capacity and demand elasticity at the same scale, the relative magnitudes of the first and second vectors are quantified, and the mismatch rate at each time scale is output. The inertia amplitude of physical nodes and the fractal elasticity demand core in time scale The degree of deviation.

[0097] The synchronization offset value is calculated by taking the natural logarithm of the mismatch rate and then taking its absolute value, as follows:

[0098]

[0099] in, This represents the synchronization offset value.

[0100] The ratio is mapped to a symmetric metric using logarithms, while the absolute value is taken so that both large and small deviations are quantified equally, generating the final synchronization offset value, which can be used directly as a scalar index in optimization constraints or penalty terms.

[0101] Construct a synchronization offset matrix based on the synchronization offset values ​​of N physical nodes. .

[0102] It should be noted that the synchronization offset matrix It is a three-dimensional matrix with a size of M×N×T.

[0103] In one embodiment of the present invention, an objective function is constructed based on the unit cost of energy storage and the unit cost of response, combined with the synchronization offset matrix, including:

[0104] Load and generate the initial energy storage capacity planning quantity and the initial response depth planning quantity for the nth physical node;

[0105] For the nth physical node, the energy storage cost is calculated based on the initial planned energy storage capacity and the unit cost of energy storage.

[0106] To convert the decision on energy storage capacity into economic value, it must be multiplied by the corresponding unit cost of energy storage. For the nth physical node, based on the initial planned energy storage capacity and the unit cost of energy storage... The energy storage cost is calculated as follows:

[0107]

[0108] in, This represents the energy storage cost of the nth physical node. This indicates the initial planned energy storage capacity.

[0109] For the nth physical node, the response cost is calculated based on the initial response depth planning quantity and the response unit cost;

[0110] Mapping the planning depth of demand response to response cost ensures the economics of the response strategy are measurable. For the nth physical node, the response cost is calculated based on the initial response depth planning amount and the response unit cost, as follows:

[0111]

[0112] in, This represents the response cost of the nth physical node. This represents the initial response depth planning quantity.

[0113] The synchronization offset matrix is ​​reduced in dimensionality according to the physical node dimension for each time scale. And every moment Sum the offsets to obtain the offset aggregate value;

[0114] The energy storage cost and response cost of each physical node are combined to form the total economic cost item; based on the preset penalty factor... The total penalty cost is calculated by combining the offset aggregate value;

[0115] Combining the total economic cost term and the total penalty cost term yields the objective function, as follows:

[0116]

[0117] in, Describe the objective function. Indicates the penalty factor .

[0118] It should be noted that the unit cost of energy storage refers to the total cost required to deploy or expand 1MWh of energy storage equipment, which typically includes: equipment investment cost, construction and installation cost, and operation and maintenance depreciation cost.

[0119] The unit cost of response refers to the cost required to achieve a demand response depth of 1MW, which typically includes: incentive compensation fees and operating and management fees.

[0120] It should be noted that the initial energy storage capacity planning amount refers to the size of the energy storage equipment capacity required to be deployed on the nth physical node in the planning stage, which reflects the initial configuration scale of the available energy storage resources for the nth physical node in subsequent scheduling.

[0121] The initial response depth planning quantity refers to the size of the demand response depth that the nth physical node can provide in advance during the planning stage, reflecting the initial design range of the load that can be adjusted for that node in subsequent scheduling.

[0122] In one embodiment of the present invention, the synchronization offset matrix is ​​reduced in dimensionality according to the dimension of the physical nodes, for each time scale. And every moment Sum the offsets to obtain the offset aggregate value, including:

[0123] The synchronization offset matrix is ​​merged according to the dimension of the physical nodes to obtain the sum of the offsets at the time scale and at each moment; where the dimension of the physical nodes is N.

[0124] It should be noted that all nodes need to be on the same time scale. Only by aggregating the offset information at time t can the synchronization mismatch status of the entire network be reflected. This can be expressed by a two-dimensional function: the sum of offsets. The sum of offsets is reflected at each time scale. The cumulative value of the offset of N nodes at time t.

[0125] The sum of the time scale and the offset at each moment is normalized according to the number of physical nodes to obtain the average offset.

[0126] To eliminate the absolute value bias caused by varying numbers of nodes N, the cumulative value needs to be standardized to generate an average offset (the ratio of the sum of offsets to N), which represents the average offset at each time scale. The average mismatch degree of a single physical node at time t.

[0127] Compare the average offset with a preset offset threshold to generate an offset aggregate value, which can be 1 or 0.

[0128] It should be noted that the continuous average offset needs to be converted into a binary indicator to quickly identify out-of-limit or normal states. The offset aggregate value is recorded as 1 when the average offset exceeds a preset offset threshold, and 0 otherwise, for subsequent decision-making or alarm purposes.

[0129] In one embodiment of the present invention, determining the energy storage configuration and demand response strategy for each physical node includes:

[0130] Determine the first and second value boundaries corresponding to the planned energy storage capacity and response depth for each physical node, respectively.

[0131] It should be noted that the first and second value boundaries clearly define the upper and lower limits of each decision variable, establish the value range of the energy storage capacity planning quantity and the response depth planning quantity, and limit the feasible domain.

[0132] Set a stopping threshold for iterating the objective function value;

[0133] If the initial response depth planning quantity and the objective function value obtained from the initial response depth planning quantity for N physical nodes are less than or equal to the iteration stopping threshold, then the initial response depth planning quantity and the initial response depth planning quantity for N physical nodes are output as the energy storage configuration and demand response strategy for the corresponding physical nodes, respectively.

[0134] Otherwise, update the initial response depth planning quantity and the initial response depth planning quantity within the first value boundary and the second value boundary until the objective function value is less than or equal to the iteration stopping threshold, and output the corresponding updated response depth planning quantity and the updated response depth planning quantity as the energy storage configuration and demand response strategy of the corresponding physical node, respectively.

[0135] It should be noted that the initial response depth planning quantity and the initial response depth planning quantity are updated based on the gradient descent method.

[0136] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A power system source-grid-load-storage coordinated intelligent planning system, characterized in that, include: The first acquisition module is used to acquire the grid load value of the target power grid at fixed time intervals; The first processing module performs discrete wavelet transform on the time-series grid load values ​​to obtain wavelet coefficients at M time scales. These coefficients are then combined with a preset fractal order to construct the corresponding fractal elastic demand kernel, including: A load sequence is constructed based on the time-series grid load values, and each grid load value in the load sequence is normalized by maximum and minimum values ​​to form a net load sequence. Set M time scales , satisfy: ,and ;in, Represents a fixed time interval, for each time scale The wavelet coefficients are obtained by calculating the discrete wavelet transform. , It is a positive integer; For each time scale The absolute values ​​of the corresponding wavelet coefficients are calculated. ; Based on fractal order and time scale Calculate the corresponding scale weights , ; For each time scale , absolute value With scale weight According to fractal order Coupling forms fractal elastic fundamental quantities; Globally normalize the fractal elasticity basic quantities to output the fractal elasticity demand kernel; The second acquisition module is used to acquire the energy storage value of each energy storage device in the nth physical node at fixed time intervals, as well as the phase angle of each energy storage device, and to determine the inertia vector phase field of the nth physical node by combining it with a preset angular frequency; 1≤n≤N, where n is a positive integer, including: Based on the energy storage value and phase angle of the energy storage device, the inertia vector component of the corresponding energy storage device is calculated in combination with the preset angular frequency. The inertia vector phase field of the nth physical node is obtained by summing the inertia vector components of all energy storage devices within the nth physical node. The second processing module is used to construct a synchronization offset matrix based on the inertia vector phase field of each physical node and the fractal elasticity requirement kernel of each time scale, including: Regarding the first At each time point, determine the inertia magnitude of the nth physical node, and the inertia magnitude at each time scale. The fractal elastic demand kernel; where the inertia amplitude of the nth physical node is the magnitude of the corresponding inertia vector phase field; The inertia magnitude of the nth physical node is expanded into a first vector of dimension M through M copies; Based on M time scales The fractal elastic demand kernel is used to construct a second vector of dimension M; Calculate the element-wise ratio of the first vector to the second vector to obtain the mismatch rate; The mismatch rate is calculated using the natural logarithm and its absolute value is taken to form the synchronization offset value. Construct a synchronization offset matrix based on the synchronization offset values ​​of the timing of N physical nodes; The intelligent planning module is used to construct an objective function based on the unit cost of energy storage and the unit cost of response, combined with the synchronization offset matrix, to determine the energy storage configuration and demand response strategy for each physical node. The specific process of constructing the objective function is as follows: Load and generate the initial energy storage capacity planning quantity and the initial response depth planning quantity for the nth physical node; For the nth physical node, the energy storage cost is calculated based on the initial planned energy storage capacity and the unit cost of energy storage. For the nth physical node, the response cost is calculated based on the initial response depth planning quantity and the response unit cost; The synchronization offset matrix is ​​reduced in dimensionality according to the physical node dimension for each time scale. And every moment Sum the offsets to obtain the offset aggregate value; The energy storage cost and response cost of each physical node are combined to form the total economic cost item; based on the preset penalty factor... The total penalty cost is calculated by combining the offset aggregate value; The objective function is obtained by combining the total economic cost term and the total penalty cost term.

2. The power system source-grid-load-storage coordinated intelligent planning system according to claim 1, characterized in that, The synchronization offset matrix is ​​reduced in dimensionality according to the physical node dimension for each time scale. And every moment Sum the offsets to obtain the offset aggregate value, including: The synchronization offset matrix is ​​merged according to the dimension of the physical nodes to obtain the sum of the offsets at the time scale and at each moment; where the dimension of the physical nodes is N. The sum of the time scale and the offset at each moment is normalized according to the number of physical nodes to obtain the average offset. Compare the average offset with a preset offset threshold to generate an offset aggregate value, which can be 1 or 0.

3. The power system source-grid-load-storage coordinated intelligent planning system according to claim 2, characterized in that, Determine the energy storage configuration and demand response strategy for each physical node, including: Determine the first and second value boundaries corresponding to the planned energy storage capacity and response depth for each physical node, respectively. Set a stopping threshold for iterating the objective function value; If the initial response depth planning quantity and the objective function value obtained from the initial response depth planning quantity for N physical nodes are less than or equal to the iteration stopping threshold, then the initial response depth planning quantity and the initial response depth planning quantity for N physical nodes are output as the energy storage configuration and demand response strategy for the corresponding physical nodes, respectively. Otherwise, update the initial response depth planning quantity and the initial response depth planning quantity within the first value boundary and the second value boundary until the objective function value is less than or equal to the iteration stopping threshold, and output the corresponding updated response depth planning quantity and the updated response depth planning quantity as the energy storage configuration and demand response strategy of the corresponding physical node, respectively.

Images