User side energy storage method for improving flexibility of power grid

Abstract

The invention discloses a user side energy storage method for improving the flexibility of a power grid, and relates to the technical field of intelligent power grids and energy storage, and the method comprises the steps: collecting the power load data and electricity price data of user side power equipment, building a load prediction model according to the power load data, and obtaining a power load prediction value; calculating a charging and discharging time window of the energy storage system based on the electricity price data and the electricity load prediction value; generating an optimal scheduling instruction of the energy storage system according to the charging and discharging time window in combination with the capacity constraint of the energy storage system; and controlling the energy storage system to perform charging and discharging operation according to the optimal scheduling instruction. Through intelligent prediction and accurate scheduling, the energy storage system can discharge for peak clipping in the peak period of the load and charge for valley filling in the valley period of the load, load fluctuation is effectively stabilized, the peak regulation capability of the power grid is enhanced, the energy storage system can quickly respond to the fluctuation of the power grid through real-time monitoring in cooperation with dynamic regulation of the ramp rate and SOC constraint, and the energy storage efficiency is improved. And flexible power support is provided.

Description

Technical Field

[0001] The present invention relates to the field of smart grid and energy storage technology, and in particular to a user-side energy storage method for improving the flexibility of a grid. Background Art

[0002] With the rapid advancement of renewable energy technology, the power system faces an increasing demand for flexibility. User-side energy storage technology has emerged to effectively regulate power supply and demand by deploying energy storage equipment at the user end. At present, the power grid not only has to cope with the increasingly fluctuating power demand, but also needs to adapt to the intermittent power generation characteristics of renewable energy sources such as wind and solar energy. The driving force behind this transformation comes from the dual pursuit of power system reliability and economy. Energy storage technology, as a key means to improve the flexibility of the power grid, has received widespread attention. In recent years, the development of user-side energy storage technology has also been accelerating, but its implementation still faces many challenges.

[0003] There are several significant deficiencies in existing user-side energy storage technologies. First, load forecasting models often rely too much on historical electricity consumption data and fail to fully consider changes in user electricity consumption habits, climate factors, and seasonal fluctuations. This single data dependence makes the forecast results susceptible to changes in the external environment, resulting in insufficient forecast accuracy, which in turn affects the dispatch efficiency of the energy storage system. According to research, the mean absolute percentage error (MAPE) of some load forecasting models is above 15%, which significantly reduces the economy and benefits of the energy storage system. Secondly, the existing technology lacks real-time response to electricity price fluctuations during the charging and discharging scheduling process. Although some systems are able to monitor changes in electricity prices, scheduling decisions are often delayed, failing to achieve timely charging during low electricity prices and effective discharge during high electricity prices, which results in users being unable to fully utilize electricity price differences to maximize their benefits and reduce economic benefits. Finally, the dispatch instructions of many energy storage systems fail to reasonably consider the capacity limitations of energy storage equipment, resulting in overload operation or energy loss. For example, some systems cannot formulate reasonable charging and discharging strategies based on the actual carrying capacity of the power grid and energy storage capacity during peak load periods, which can easily lead to equipment failure or reduced efficiency. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention proposes a user-side energy storage method for improving the flexibility of the power grid.

[0005] Therefore, the present invention can solve the problems mentioned in the background art.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a user-side energy storage method for improving the flexibility of a power grid, which includes:

[0008] Collecting power load data and electricity price data of power equipment on the user side, establishing a load forecasting model based on the power load data, and obtaining a power load forecast value;

[0009] Based on the electricity price data and the electricity load forecast value, the charging and discharging time window of the energy storage system is calculated:

[0010] According to the charging and discharging time window and in combination with the capacity constraint of the energy storage system, an optimization scheduling instruction of the energy storage system is generated;

[0011] The optimized scheduling instruction is sent to the power converter of the energy storage system, and the energy storage system is controlled to perform charging and discharging operations according to the optimized scheduling instruction.

[0012] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid described in the present invention, wherein: the power load data includes active power, reactive power, power factor, voltage and current values ​​in each time period;

[0013] The electricity price data includes peak and valley time-of-use electricity prices, peak electricity prices and valley bottom electricity prices;

[0014] An enhanced feature vector is constructed based on the power load data; the enhanced feature vector includes a product term of the load sequence and the time weight, a power consumption feature matrix, and a volatility index.

[0015] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid described in the present invention, wherein: the load forecasting model includes a first calculation branch, a second calculation branch, a memory unit and a residual connection term;

[0016] The first calculation branch receives the enhanced feature vector input and combines it with the memory unit output, and obtains a first hidden state through a nonlinear activation function process;

[0017] The second calculation branch receives the enhanced feature vector input and combines it with the residual connection term to obtain a second hidden state through a nonlinear activation function process;

[0018] The first hidden state and the second hidden state are combined by using a fusion function, and the prediction result is corrected according to a dynamically adjusted adaptive correction coefficient to obtain the power load prediction value.

[0019] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid described in the present invention, wherein: the load forecasting model also includes optimizing the load forecasting model by using a loss function with a smoothing constraint;

[0020] The smoothness constraint is obtained by calculating the difference of the prediction values ​​at adjacent moments and applying a time decay weight, as shown in the following formula:

[0021] S(t)=∑|ΔY(t)|·exp(-δt)

[0022] The loss function is shown as follows:

[0023]

[0024] in, is the actual load value at time t, Y(t) is the predicted load value at time t, S(t) is the smoothing constraint term, λ is the weight coefficient, δ is the time attenuation factor, L * (t) is the loss function, and ΔY(t) is the difference between the predicted values ​​at adjacent moments.

[0025] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid described in the present invention, the calculation of the charging and discharging time window includes:

[0026] Divide 24 hours into N time segments using the sliding segmentation method;

[0027] Based on the daily average electricity price P avg Calculate peak and trough determination thresholds;

[0028] Calculate the load rate for each time segment rate (t), the load rate is the ratio of the predicted power load value to the historical maximum load;

[0029] Among them, the charging time corresponds to the low electricity price period, and the discharging time corresponds to the peak electricity price period;

[0030] During the low electricity price period, when the electricity price is lower than the low valley threshold and the load rate is lower than the first preset value, three or more consecutive time segments are selected to form a charging time window. The charging window determination condition is expressed as w charge (t) = 1;

[0031] During the peak period of electricity price, when the electricity price is higher than the peak threshold and the load rate is higher than the second preset value, two or more consecutive time segments are selected to form a discharge time window. The discharge window determination condition is expressed as W discharge (t)=1.

[0032] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid described in the present invention, generating an optimization dispatch instruction for the energy storage system includes:

[0033] According to the obtained charge and discharge time window mark W charge (t) and W discharge (t), divide a day into multiple charging intervals and discharging intervals;

[0034] Calculating the target charge and discharge power of the time segment t of the charging interval and the discharging interval respectively, and setting constraints;

[0035] Calculating the charging and discharging duration of the time segments t of the charging interval and the discharging interval respectively;

[0036] The calculated charging and discharging power sequence of each time segment and the corresponding duration are combined to form a scheduling instruction set;

[0037] Each optimization scheduling instruction in the scheduling instruction set includes a start and end time, a power value and a running time.

[0038] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid according to the present invention, the target charging power for the time segments t of the charging interval and the discharging interval are calculated respectively, including:

[0039] When W charge When (t) = 1:

[0040] The target charging power is calculated as follows:

[0041]

[0042] Among them, P c (t) is the target charging power, P c，max is the maximum charging power, SOC max is the maximum state of charge allowed by the energy storage system, SOC(t-1) is the state of charge at time segment t-1, and E rated is the rated capacity, η c is the charging efficiency, Δt is the length of each time segment;

[0043] When W discharge When (t) = 1:

[0044] The target discharge power is calculated as follows:

[0045]

[0046] Among them, P c (t) is the target discharge power, P d，max is the maximum discharge power, η d is the discharge efficiency, SOC min It is the minimum state of charge allowed by the energy storage system.

[0047] As a preferred solution of the user-side energy storage method for improving the flexibility of the power grid according to the present invention, the charging and discharging time includes:

[0048] The charging time is calculated based on the target SOC increment and charging power integral of each time segment in the charging interval, as shown in the following formula:

[0049] T c =∑(Δt·W charge (t))

[0050] The discharge duration is calculated based on the target SOC reduction and discharge power integral of each time segment in the discharge interval, as shown in the following formula:

[0051] T d =∑(Δt·W discharge (t))

[0052] Among them, T c is the charging time, T d is the discharge time.

[0053] In a second aspect, an embodiment of the present invention provides a user-side energy storage system for improving the flexibility of a power grid, which includes:

[0054] A data collection module is used to collect power load data and electricity price data of power equipment on the user side;

[0055] An electricity load prediction module is used to establish a load prediction model according to the electricity load data to obtain an electricity load prediction value;

[0056] A charging and discharging time window calculation module, used to calculate the charging and discharging time window of the energy storage system based on the electricity price data and the electricity load forecast value;

[0057] The scheduling instruction generation module is used to generate an optimized scheduling instruction for the energy storage system based on the charging and discharging time window and in combination with the capacity constraint of the energy storage system; and control the energy storage system to perform charging and discharging operations according to the optimized scheduling instruction.

[0058] The beneficial effect of the present invention is to obtain power load and electricity price data through high-precision acquisition equipment, and then accurately predict the power load using the prediction model constructed by the dual-stream neural network, determine the optimal charging and discharging time window based on the prediction results combined with the time-of-use electricity price, and finally generate detailed scheduling instructions according to the various constraints of the energy storage system, so as to achieve the economic scheduling goal of storing energy during the low electricity price period and discharging during the peak period, effectively reducing the electricity cost while improving the stability of the power grid operation. Through intelligent prediction and precise scheduling, the energy storage system can discharge and shave the peak during the peak load period and charge and fill the valley during the valley period, effectively smoothing the load fluctuation and enhancing the peak-shaving capacity of the power grid; at the same time, the system monitors the load rate and power factor in real time, and cooperates with the dynamic adjustment of the climbing rate and SOC constraints, so that the energy storage system can quickly respond to power grid fluctuations, provide flexible power support, and enhance the dynamic adjustment capability of the power grid; this active control mode based on prediction significantly improves the operational flexibility of the power grid compared with the traditional passive response method. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0060] Figure 1 Flowchart of a user-side energy storage method to improve grid flexibility. DETAILED DESCRIPTION

[0061] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0062] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0063] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0064] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0065] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0066] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0067] Example 1

[0068] Reference Figure 1 , which is the first embodiment of the present invention, and provides a user-side energy storage method for improving the flexibility of a power grid, comprising:

[0069] S1: Collecting power load data and electricity price data of power equipment on the user side, and transmitting the power load data and the electricity price data to a control unit through a data acquisition device;

[0070] Collect power load data of power equipment on the user side, wherein the power load data includes: active power, reactive power, power factor, voltage and current values ​​in each time period; collect electricity price data of power grid operation, wherein the electricity price data includes: peak-valley time-of-use electricity price, peak electricity price and valley electricity price; the data acquisition device is provided with an electric energy metering module and a communication module, and the power load data on the user side is recorded in real time through the electric energy metering module, and the sampling period is 15 minutes; the communication module establishes a connection with the power grid dispatching system using RS485 communication method to obtain the electricity price data; the power load data and the electricity price data are transmitted to the data processing module of the control unit through the communication module, and the data processing module performs standardization processing and data quality verification on the received data.

[0071] S2: Establishing a load prediction model according to the power load data;

[0072] A load forecasting model is established according to the power load data. The specific construction process of the load forecasting model includes: dividing the power load data into a training data set and a verification data set according to the time series, the training data set accounts for 80%, and the verification data set accounts for 20%; based on the power load data, an enhanced feature vector is constructed by multiple feature extraction, the enhanced feature vector includes three components: the product term of the load sequence and the time weight, the power feature matrix, and the volatility index, and a dynamic adjustment factor is introduced to adaptively adjust the feature vector. The calculation formula can be expressed as follows:

[0073] X(t)=[L(t)·W(t), E(t), V(t)]·θ(t)

[0074] Among them, L(t) represents the load sequence, W(t) is the time weight coefficient, E(t) is the power consumption characteristic matrix, V(t) is the volatility index, and θ(t) is the dynamic adjustment factor.

[0075] The load forecasting model adopts a dual-stream neural network structure. The first calculation branch receives the enhanced feature vector input and combines it with the memory unit output. The second calculation branch also receives the enhanced feature vector input and introduces a residual connection term. The outputs of the two branches are processed by a nonlinear activation function to form two hidden states. The core calculation process can be expressed as follows:

[0076] H1(t)=σ(W1·X(t)+α·M(t))

[0077] H2(t)=σ(W2·X(t)+β·R(t))

[0078] Y(t)=φ(H1(t),H2(t))·K(t)

[0079] Among them, H1(t) and H2(t) are the hidden states of the two parallel branches, M(t) is the memory unit output, R(t) is the residual connection term, K(t) is the adaptive correction coefficient, α and β are balance parameters, Y(t) is the power load prediction value, W1 and W2 are the weight matrices of the neural network, φ() is the fusion function used to merge the outputs of the two branches, and σ is the activation function;

[0080] The two hidden states are then merged through a fusion function, and the prediction result is corrected by an adaptive correction coefficient, wherein the adaptive correction coefficient is dynamically adjusted according to the prediction error; the prediction model is optimized using a loss function with a smooth constraint, and the smooth constraint is obtained by calculating the difference of the prediction values ​​at adjacent moments and applying a time decay weight, which is used to suppress the violent fluctuation of the prediction result, as shown in the following formula:

[0081]

[0082] S(t)=∑|ΔY(t)|·exp(-δt)

[0083] in, is the actual load value, S(t) is the smoothness constraint term, λ is the weight coefficient, δ is the time attenuation factor, L*(t) is the loss function, and ΔY(t) is the difference between adjacent moments of the predicted value.

[0084] Finally, the load forecasting model is trained using the Adam optimization algorithm, and the number of training rounds is set to 200.

[0085] It should be noted that the acquisition of the L(t) load sequence extracts the historical load values ​​of the last 168 hours from the power load data to form an initial sequence, and the initial sequence is normalized, and the normalization interval is set to [-1, 1]; the load sequence is updated by a sliding time window, and the window length is set to 168 hours;

[0086] The calculation of the W(t) time weight coefficient is based on the time correlation of the load sequence, and the weight coefficient is calculated using an exponential decay function. The data closer to the prediction time is given a larger weight value. The weight coefficient takes a value in the interval [0.6, 1], and the decay rate is set to 0.05;

[0087] The construction of the E(t) electricity consumption characteristic matrix includes segmenting the electricity load data according to a 24-hour period, extracting statistical features such as the maximum value, minimum value, mean value, standard deviation, etc. in each segment, and forming a 4×24-dimensional characteristic matrix. The characteristic matrix is ​​updated every 24 hours;

[0088] The determination of the V(t) volatility index is based on the first-order difference value of the load sequence, the load change rate at adjacent moments is calculated, and the volatility index is obtained by exponential smoothing method, the smoothing coefficient is set to 0.3, and the volatility index is updated once every hour;

[0089] The generation of the θ(t) dynamic adjustment factor is based on the change trend of the load forecast error, and the adjustment factor is calculated using an adaptive algorithm. When the forecast error increases, the adjustment factor decreases accordingly, and the value range is limited to [0.8, 1.2];

[0090] The output of the M(t) memory unit is obtained by recursively processing the load sequence characteristics. The memory unit adopts a gated structure, including three control units: an input gate, a forget gate, and an output gate. The threshold values ​​are all set to 0.5.

[0091] The calculation of the R(t) residual connection term performs a difference operation between the input feature and the output of the previous layer network, and performs feature mapping through a 1×1 convolution kernel to obtain a residual connection term for maintaining low-level feature information;

[0092] The updating of the K(t) adaptive correction coefficient is based on the statistical distribution of the prediction error. The correction coefficient is calculated using a normalized exponential function. When the prediction error exceeds a set threshold, the correction mechanism is triggered. The value range of the correction coefficient is [0.9, 1.1].

[0093] S3: Based on the electricity price data and the electricity load forecast value, calculate the charging and discharging time window of the energy storage system, wherein the charging time corresponds to the low electricity price period and the discharging time corresponds to the peak electricity price period;

[0094] Based on the electricity price data and the electricity load forecast value, a charging and discharging time window of the energy storage system is calculated, and the calculation process includes:

[0095] The sliding segmentation method is used to divide 24 hours into N time segments, each of which is 30 minutes long. Based on the daily average electricity price P avg Calculate the peak and valley determination thresholds, the peak threshold P high is 1.2 times the daily average electricity price, and the valley threshold Pl ow =0.85 times the daily average electricity price; calculate the load factor Load for each time segment rate (t), the load rate is the ratio of the predicted load value to the historical maximum load.

[0096] During the valley period, when the electricity price is lower than the valley threshold and the load rate is lower than 0.7, three or more consecutive time segments are selected to form a charging time window. The charging window determination condition is expressed as W charge (t)=1.

[0097] During the peak period, when the electricity price is higher than the peak threshold and the load factor is higher than 0.85, two or more consecutive time segments are selected to form a discharge time window. The discharge window determination condition is expressed as W discharge (t)=1.

[0098] The time interval Δt between the charging time window and the discharging time window is not less than 2 hours;

[0099] It should be noted that the electricity price data is recorded as P(t), which represents the value of electricity at time t; the daily average electricity price P is calculated avg :

[0100]

[0101] The charging window mark W charge The calculation process of (t) is:

[0102] W charge (t) = AND (P flag (t), L flag (t), C flag (t))

[0103] Among them, when P(t) <P low When P flag (t) = 1;

[0104] When Load rate When (t)<0.7, L flag (t) = 1;

[0105] When the number of time segments that continuously meet the above conditions is ≥ 3, C flag (t) = 1;

[0106] The discharge window mark W discharge The calculation process of (t) is:

[0107]

[0108] in:

[0109] When P(t)>P high hour,

[0110] When Load rate When (t)>0.85,

[0111] When the number of time segments that continuously meet the above conditions is ≥ 2, D flag (t) = 1;

[0112] Among them, P flag (t) is the low electricity price mark, L flag (t) is the low load factor mark, C flag (t) is the charging continuity mark, It is the peak mark of electricity price. D is the high load factor indicator. flag (t) is the discharge continuity mark.

[0113] S4: generating an optimization scheduling instruction for the energy storage system according to the charging and discharging time window and in combination with the capacity constraint of the energy storage system, wherein the optimization scheduling instruction includes the charging and discharging power and the charging and discharging duration;

[0114] First, based on the 48 time segments (30 minutes each) obtained by sliding segmentation, the charge and discharge time window marker W charge (t) and W discharge (t), a day is divided into multiple charging intervals and discharging intervals. charge The continuous time segment with (t) = 1 is marked as the charging interval. discharge The continuous time segments with (t)=1 are marked as discharge intervals.

[0115] Then set the constraints of the energy storage system, including the rated capacity E rated , Maximum charging power P c，max , Maximum discharge power P d，max And the state of charge SOC safe operating range [SOC min , SOC max ], where SOC minSet to 0.1, SOC max Set to 0.9, the charge and discharge efficiencies are η c and η d , the power ramp rate is limited to δP max Specific constraints include:

[0116] SOC min ≤SOC(t)≤SOC max

[0117] -P d，max ≤P(t)≤P c，max

[0118] |P(t)-P(t-1)|≤δP max

[0119] It should be noted that the charge state evolution equation is as follows:

[0120]

[0121] Where t is the time segment within the charging interval.

[0122] Furthermore, in the time segment t within the charging interval, based on the load rate Load rate (t) and electricity price P(t) to calculate the target charging power P c (t), when the load rate Load rate (t)<0.7 and electricity price P(t)<P low When W is selected, the charging power must meet the maximum power limit, SOC upper limit constraint and climbing rate limit at the same time. charge When (t) = 1:

[0123] Power calculation:

[0124]

[0125] Load rate constraints:

[0126] if Load rate (t)<0.7: P c (t) = α·P c，max

[0127] Ramp rate limit:

[0128] |P c (t)-P c (t-1)|≤δP max

[0129] Furthermore, in the time segment t within the discharge interval, based on the load rate Loadrate (t) and electricity price P(t) to calculate the target discharge power P d (t), when the load rate Load rate (t)>0.85 and electricity price P(t)>P high When W is selected, the discharge power must meet the maximum power limit, SOC lower limit constraint and climbing rate limit at the same time. discharge When (t) = 1:

[0130] Power calculation:

[0131]

[0132] Load rate constraints:

[0133] if Load rate (t)>0.85:P d (t) = β·P d，max

[0134] Hill climbing constraints:

[0135] |P d (t)-P d (t-1)|≤δP max

[0136] Further, the charging time and the discharging time are calculated, wherein the charging time T c The discharge time T is calculated based on the target SOC increment and charging power integral of each time segment in the charging interval. d It is calculated based on the target SOC reduction and discharge power integration of each time segment in the discharge interval, and the charge and discharge interval must meet the minimum number of consecutive time periods. Specific calculation:

[0137] T c =∑(Δt·W charge (t))

[0138] T d =∑(Δt·W discharge (t))

[0139] Its charge and discharge energy balance constraints:

[0140]

[0141] Finally, the calculated charging and discharging power sequences and corresponding durations of each time segment are combined to form a scheduling instruction set. Each instruction contains the start and end time, power value and operating duration. The interval between adjacent charging and discharging intervals is not less than 4 time segments (i.e. 2 hours). A ramp transition instruction is generated for the power change point, and an appropriate power adjustment margin is reserved.

[0142] The optimized scheduling instruction is sent to the power converter of the energy storage system, and the energy storage system is controlled to perform charging and discharging operations according to the optimized scheduling instruction.

[0143] Furthermore, this embodiment also provides a user-side energy storage system for improving the flexibility of the power grid, including:

[0144] A data collection module is used to collect power load data and electricity price data of power equipment on the user side;

[0145] An electricity load prediction module is used to establish a load prediction model according to the electricity load data to obtain an electricity load prediction value;

[0146] A charging and discharging time window calculation module, used to calculate the charging and discharging time window of the energy storage system based on the electricity price data and the electricity load forecast value;

[0147] The scheduling instruction generation module is used to generate an optimized scheduling instruction for the energy storage system based on the charging and discharging time window and in combination with the capacity constraint of the energy storage system; and control the energy storage system to perform charging and discharging operations according to the optimized scheduling instruction.

[0148] This embodiment also provides a computer device, which is suitable for the user-side energy storage method for improving the flexibility of the power grid, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the user-side energy storage method for improving the flexibility of the power grid as proposed in the above embodiment.

[0149] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0150] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the user-side energy storage method for improving the flexibility of the power grid as proposed in the above embodiment is implemented.

[0151] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be found in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0152] Example 2

[0153] This is the second embodiment of the present invention, which provides a user-side energy storage method for improving the flexibility of the power grid. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0154] This test was conducted in an industrial park, and a 10MW / 40MWh lithium-ion battery energy storage power station in the park was selected as the test object. Before the test, the energy storage system was fully tested to ensure that all performance parameters met the requirements. The data acquisition system uses Schneider PM5560 high-precision energy meter and ABB RTU 560 data acquisition terminal to transmit data to the data processing server in the control room through the RS485 bus. The sampling period is set to 15 minutes, and 30 days of historical operation data are continuously collected as the basis for model training.

[0155] In order to verify the optimization scheduling effect of the present invention, the park was divided into two areas, A and B, for comparative tests. Area A adopted the traditional fixed-time scheduling strategy, and area B adopted the intelligent optimization scheduling scheme of the present invention. During the test, the power load data of the two areas were recorded in real time by the data acquisition device, including active power, reactive power, power factor, voltage and current values ​​at 15-minute intervals. At the same time, real-time electricity price data was obtained from the power grid dispatching system, including peak and valley time-of-use electricity prices, peak electricity prices and valley bottom electricity prices.

[0156] The data processing module performs standardization and quality verification on the collected data, and inputs the data into the load forecasting model after removing outliers. The forecasting model adopts the dual-stream neural network structure proposed in this invention. Based on 168 hours of historical data, it combines time weight, power consumption feature matrix and volatility index to construct an enhanced feature vector. The model training adopts the Adam optimization algorithm, sets the learning rate to 0.001, and the training rounds to 200. The forecast results are compared with the actual load to verify the forecast accuracy.

[0157] Based on the predicted load and electricity price data, the system automatically calculates the charging and discharging time window of the energy storage system. The optimal charging and discharging interval is determined by sliding segmentation in units of 30 minutes, combined with the peak threshold of 1.2 times the average daily electricity price and the valley threshold of 0.85 times the average daily electricity price. Under the conditions of meeting the SOC constraints [0.1, 0.9] and the power ramp rate limit of ±20% rated power / minute, detailed charging and discharging scheduling instructions are generated. The specific experimental data are shown in Table 1 below:

[0158] Table 1 Experimental data table

[0159]

[0160]

[0161] By comparing the test data, it can be seen that the optimized scheduling scheme (area B) of the present invention has significant advantages over the traditional fixed time scheduling strategy (area A):

[0162] The average load factor in Area B increased to 82.3%, 4.8 percentage points higher than that in Area A, indicating that the optimized dispatching scheme can better smooth out load fluctuations. The peak-to-valley difference increased from 4.2MW to 6.8MW, an increase of 61.9%, indicating that the system has a stronger peak-shaving capability.

[0163] The electricity price sensitive response time was reduced from 25 minutes to 12 minutes, an increase of 52%, which reflects the advantages of the present invention in terms of market response speed. This is due to the synergy of the prediction model and dynamic adjustment mechanism of the dual-stream neural network structure.

[0164] The daily revenue under the optimized scheduling scheme reached 4,260 yuan, an increase of 49.5% over the traditional scheme. The charge-discharge cycle efficiency increased by 4.1 percentage points to 88.7%, mainly due to more accurate charge-discharge time window selection and power optimization control.

[0165] The power factor of area B is increased to 0.95, and the power quality improvement rate is 15.3%, which is 3.3% and 80% higher than that of area A. This shows that the present invention not only optimizes the economic benefits, but also effectively improves the power quality indicators.

[0166] To verify the stability and reliability of the solution, parallel tests were conducted in four areas, C, D, E, and F. The data showed that the performance indicators in each area had a small fluctuation range: the average load rate was between 81.9% and 83.1%, the response time was between 11 and 13 minutes, the daily income was between 4180 and 4380 yuan, and the charge and discharge efficiency was between 88.5% and 89.1%. This stable performance shows that the solution of the present invention has good adaptability and scalability.

[0167] In summary, the energy storage optimization scheduling scheme proposed in this invention is significantly superior to traditional schemes in terms of load regulation capability, market response speed, economic benefits and power quality, and has good stability and reliability. This scheduling strategy based on intelligent prediction and multi-constraint optimization provides a new technical path for the efficient utilization of energy storage systems.

[0168] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A user-side energy storage method for improving power grid flexibility, characterized in that: include: Collecting power load data and electricity price data of power equipment on the user side, establishing a load forecasting model based on the power load data, and obtaining a power load forecast value; Calculating a charging and discharging time window of an energy storage system based on the electricity price data and the electricity load forecast value; According to the charging and discharging time window and in combination with the capacity constraint of the energy storage system, an optimization scheduling instruction of the energy storage system is generated; The optimized scheduling instruction is sent to the power converter of the energy storage system, and the energy storage system is controlled to perform charging and discharging operations according to the optimized scheduling instruction.

2. The user-side energy storage method for improving power grid flexibility according to claim 1, characterized in that: The power load data includes active power, reactive power, power factor, voltage and current values ​​in each time period; The electricity price data includes peak and valley time-of-use electricity prices, peak electricity prices and valley bottom electricity prices; An enhanced feature vector is constructed based on the power load data; the enhanced feature vector includes a product term of the load sequence and the time weight, a power consumption feature matrix, and a volatility index.

3. The user-side energy storage method for improving power grid flexibility according to claim 2, characterized in that: The load forecasting model includes a first calculation branch, a second calculation branch, a memory unit and a residual connection term; The first calculation branch receives the enhanced feature vector input and combines it with the memory unit output, and obtains a first hidden state through a nonlinear activation function process; The second calculation branch receives the enhanced feature vector input and combines it with the residual connection term to obtain a second hidden state through a nonlinear activation function process; The first hidden state and the second hidden state are combined by using a fusion function, and the prediction result is corrected according to a dynamically adjusted adaptive correction coefficient to obtain the power load prediction value.

4. The user-side energy storage method for improving power grid flexibility according to claim 3, characterized in that: The load forecasting model also includes optimizing the load forecasting model using a loss function with a smoothness constraint; The smoothness constraint is obtained by calculating the difference of the prediction values ​​at adjacent moments and applying a time decay weight, as shown in the following formula: S(t)=∑|ΔY(t)|·exp(-δt) The loss function is shown as follows: in, is the actual load value at time t, Y(t) is the predicted load value at time t, S(t) is the smoothing constraint term, λ is the weight coefficient, δ is the time attenuation factor, L * (t) is the loss function, and ΔY(t) is the difference between the predicted values ​​at adjacent moments.

5. The user-side energy storage method for improving power grid flexibility according to claim 4, characterized in that: The calculation of the charge and discharge time window includes: Divide 24 hours into N time segments using the sliding segmentation method; Based on the daily average electricity price P avg Calculate peak and trough determination thresholds; Calculate the load rate for each time segment rate (t), the load rate is the ratio of the predicted power load value to the historical maximum load; Among them, the charging time corresponds to the low electricity price period, and the discharging time corresponds to the peak electricity price period; During the low electricity price period, when the electricity price is lower than the low valley threshold and the load rate is lower than the first preset value, three or more consecutive time segments are selected to form a charging time window. The charging window determination condition is expressed as W charge (t) = 1; During the peak period of electricity price, when the electricity price is higher than the peak threshold and the load rate is higher than the second preset value, two or more consecutive time segments are selected to form a discharge time window. The discharge window determination condition is expressed as W discharge (t)=1.

6. The user-side energy storage method for improving power grid flexibility according to claim 5, characterized in that: Generating optimized dispatch instructions for energy storage systems includes: According to the obtained charge and discharge time window mark W charge (t) and W discharge (t), divide a day into multiple charging intervals and discharging intervals; Calculating the target charge and discharge power of the time segment t of the charging interval and the discharging interval respectively, and setting constraints; Calculating the charging and discharging duration of the time segments t of the charging interval and the discharging interval respectively; The calculated charging and discharging power sequence of each time segment and the corresponding duration are combined to form a scheduling instruction set; Each optimization scheduling instruction in the scheduling instruction set includes a start and end time, a power value and a running time.

7. The user-side energy storage method for improving power grid flexibility according to claim 6, characterized in that: Calculating the target charging power for the time segment t of the charging interval and the discharging interval respectively includes: When W charge When (t) = 1: The target charging power is calculated as follows: Among them, P c (t) is the target charging power, P c,max is the maximum charging power, SOC max is the maximum state of charge allowed by the energy storage system, SOC(t-1) is the state of charge at time segment t-1, and E rated is the rated capacity, η c is the charging efficiency, Δt is the length of each time segment; When W discharge When (t) = 1: The target discharge power is calculated as follows: Among them, P c (t) is the target discharge power, P d,max is the maximum discharge power, η d is the discharge efficiency, SOC min It is the minimum state of charge allowed by the energy storage system.

8. The user-side energy storage method for improving power grid flexibility according to claim 7, characterized in that: The charging and discharging time includes: The charging time is calculated based on the target SOC increment and charging power integral of each time segment in the charging interval, as shown in the following formula: T c =∑(Δt·W charge (t)) The discharge duration is calculated based on the target SOC reduction and discharge power integral of each time segment in the discharge interval, as shown in the following formula: T d =∑(Δt·W discharge (t)) Among them, T c is the charging time, T d is the discharge time.

9. A user-side energy storage system for improving the flexibility of a power grid, based on the user-side energy storage method for improving the flexibility of a power grid according to any one of claims 1 to 8, characterized in that: include: A data collection module is used to collect power load data and electricity price data of power equipment on the user side; An electricity load prediction module is used to establish a load prediction model according to the electricity load data to obtain an electricity load prediction value; A charging and discharging time window calculation module, used to calculate the charging and discharging time window of the energy storage system based on the electricity price data and the electricity load forecast value; The scheduling instruction generation module is used to generate an optimized scheduling instruction for the energy storage system based on the charging and discharging time window and in combination with the capacity constraint of the energy storage system; and control the energy storage system to perform charging and discharging operations according to the optimized scheduling instruction.

Images