Light storage energy storage regulation and control method and system based on multi-time scale rolling optimization

Abstract

The invention provides an optical energy storage regulation and control method and system based on multi-time scale rolling optimization in the technical field of energy storage regulation and control, and the method comprises the steps: S1, inputting a 24-hour electricity price prediction curve, a 24-hour photovoltaic output prediction curve, a 24-hour load prediction curve and an initial SOC into a day-ahead prediction model, a 24-hour SOC reference trajectory and a standard charging and discharging plan are obtained; s2, collecting feedback data, and inputting the 24-hour SOC reference trajectory, the standard charging and discharging plan and the feedback data into the intra-day prediction model to obtain a corrected charging and discharging plan; and S3, setting an SOC dynamic limit value, collecting a real-time electricity price and real-time photovoltaic output, and inputting the corrected charging and discharging plan, the SOC dynamic limit value, the real-time electricity price, the real-time photovoltaic output and the actual SOC into the real-time prediction model to obtain a charging and discharging instruction and frequency modulation output power. The method has the advantages that the precision, economical efficiency and power grid adaptability of energy storage regulation and control are greatly improved, and the attenuation speed of the service life of the battery is greatly delayed.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage regulation, and particularly to a method and system for regulating energy storage of a photovoltaic energy storage system based on multi-time scale rolling optimization. Background Art

[0002] With the strong advocacy and promotion of clean energy globally, the proportion of photovoltaic power generation in the power system is increasing day by day. However, affected by natural conditions (such as light intensity, weather changes, etc.), photovoltaic power generation shows obvious intermittency and volatility, which poses severe challenges to the stable and reliable operation of the power system. As an effective means to address this problem, an energy storage system can store electrical energy when the photovoltaic output is excessive and release electrical energy when the photovoltaic output is insufficient, thereby suppressing power fluctuations and enhancing the stability of the power system. At the same time, the real-time electricity price mechanism is widely applied in the electricity market, making the electricity price fluctuate in real time according to the relationship between electricity supply and demand. Therefore, there is a need to enable the energy storage system to dynamically optimize the charge and discharge strategy based on real-time electricity price signals and photovoltaic output conditions to achieve efficient regulation. However, the current photovoltaic energy storage power stations face the following core challenges:

[0003] 1. Problem of time scale fragmentation: Traditional methods use a single time scale (such as only day-ahead optimization), which cannot take into account the medium- and long-term electricity price trends and real-time fluctuation characteristics, resulting in a large deviation between the charge and discharge strategy of the energy storage system and the actual electricity price.

[0004] 2. Problem of cumulative prediction error: The accuracy of photovoltaic output prediction and electricity price prediction is insufficient, and there is a lack of a dynamic compensation mechanism, resulting in frequent adjustment of the charge and discharge of the energy storage system and accelerating the attenuation of battery life.

[0005] 3. Problem of contradiction between economy and life: Ignoring the cost of battery life loss and overemphasizing short-term benefits at the expense of the economy of the entire life cycle of the energy storage system.

[0006] 4. Problem of insufficient grid coordination: The ability of the energy storage system to participate in auxiliary services such as frequency modulation and standby is not fully utilized, and the adaptability to the power grid is poor.

[0007] Therefore, how to provide a method and system for regulating energy storage of a photovoltaic energy storage system based on multi-time scale rolling optimization to improve the accuracy, economy and grid adaptability of energy storage regulation and delay the attenuation speed of battery life has become an urgent technical problem to be solved. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to provide a method and system for regulating energy storage of a photovoltaic energy storage system based on multi-time scale rolling optimization to improve the accuracy, economy and grid adaptability of energy storage regulation and delay the attenuation speed of battery life.

[0009] In a first aspect, the present invention provides a method for regulating and controlling a photovoltaic energy storage system based on multi-time scale rolling optimization, comprising the following steps:

[0010] Step S1: Create and train an electricity price prediction model, a photovoltaic power output prediction model, a load prediction model, a day-ahead prediction model, an intraday prediction model, and a real-time prediction model;

[0011] Step S2: Obtain historical electricity price data, historical photovoltaic power output data, historical load data, and the initial SOC of the energy storage system. Input the historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, input the historical photovoltaic power output data into the photovoltaic power output prediction model to obtain a 24-hour photovoltaic power output prediction curve, and input the historical load data into the load prediction model to obtain a 24-hour load prediction curve;

[0012] Step S3: Input the 24-hour electricity price prediction curve, the 24-hour photovoltaic power output prediction curve, the 24-hour load prediction curve, and the initial SOC into the day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a reference charge-discharge plan;

[0013] Step S4: Collect feedback data including the actual SOC, photovoltaic power output deviation, and frequency modulation response records. Input the 24-hour SOC reference trajectory, the reference charge-discharge plan, and the feedback data into the intraday prediction model to obtain a corrected charge-discharge plan;

[0014] Step S5: Set the SOC dynamic limit, collect the real-time electricity price and the real-time photovoltaic power output. Input the corrected charge-discharge plan, the SOC dynamic limit, the real-time electricity price, the real-time photovoltaic power output, and the actual SOC into the real-time prediction model to obtain a charge-discharge command and a frequency modulation output power. Perform energy storage regulation based on the charge-discharge command and the frequency modulation output power;

[0015] Step S6: Roll and optimize the outputs of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model;

[0016] Step S7: Set a deviation compensation mechanism, and compensate the frequency modulation output power based on the deviation compensation mechanism.

[0017] Further, in step S1, the electricity price prediction model is constructed based on a first LSTM network and is used to output a 24-hour electricity price prediction curve according to the historical electricity price data;

[0018] The photovoltaic power output prediction model is constructed based on an NWP meteorological network and an XGBoost network and is used to output a 24-hour photovoltaic power output prediction curve according to the historical photovoltaic power output data;

[0019] The load forecasting model is constructed based on the second LSTM network and is used to output a 24-hour load forecasting curve according to the historical load data.

[0020] Further, in step S1, the day-ahead forecasting model is used to output a 24-hour SOC reference trajectory and a reference charge-discharge plan based on a 24-hour electricity price forecasting curve, a 24-hour photovoltaic output forecasting curve, a 24-hour load forecasting curve, and an initial SOC at a resolution of 1 hour; the 24-hour SOC reference trajectory serves as the boundary condition for the intra-day forecasting model.

[0021] The objective function output by the day-ahead forecasting model is to maximize the value of the full-cycle profit function, which is constrained by a storage dynamic constraint function, a grid interaction constraint function, a charge-discharge mutual exclusion constraint function, and a photovoltaic output uncertainty robust optimization constraint function.

[0022] The formula for the full-cycle profit function is:

[0023]

[0024] where Max Profit DA represents the maximum value of the full-cycle profit of the energy storage system; λ DA (t) represents the predicted electricity price value at the t-th hour of the day-ahead; P dischar (t) represents the discharge power of the energy storage system at the t-th hour of the day-ahead; P char (t) represents the charging power of the energy storage system at the t-th hour of the day-ahead; Δt DA represents the time resolution of the day-ahead forecasting model, i.e., 1 hour; C degradation represents the life loss cost based on the Rainflow counting method; C grid represents the grid interaction penalty term, which is used to suppress the power peak-valley fluctuation.

[0025] The formula for the storage dynamic constraint function is:

[0026]

[0027] where SOC(t) represents the SOC of the energy storage system at the t-th hour of the day-ahead; SOC(t - 1) represents the SOC of the energy storage system at the (t - 1)-th hour of the day-ahead; η char represents the charging efficiency of the energy storage system; represents the discharge efficiency of the energy storage system; E rated represents the rated capacity of the energy storage system.

[0028] The formula for the grid interaction constraint function is:

[0029] P grid,min (t) ≤ P grid (t) = Pload (t) + P char (t) - P pv_pred (t) - P dischar (t) ≤

[0030] P grid,max (t);

[0031] Among them, P grid,min (t) represents the minimum grid output power at the t-th hour before the day; P grid (t) represents the grid power at the t-th hour before the day; P load (t) represents the load power at the t-th hour before the day; P pv_pred (t) represents the predicted value of the photovoltaic output at the t-th hour before the day; P grid,max (t) represents the maximum grid output power at the t-th hour before the day;

[0032] The formula of the charge-discharge mutual exclusion constraint function is:

[0033] P dischar (t) · P char (t) = 0;

[0034] The formula of the photovoltaic output uncertainty robust optimization constraint function is:

[0035] P pv_min (t) ≤ P pv_pred (t) ≤ P pv_max (t);

[0036] Among them, P pv_min (t) represents the minimum value of the predicted photovoltaic output at the t-th hour before the day; P pv_max (t) represents the maximum value of the predicted photovoltaic output at the t-th hour before the day; and P pv_min (t) = 0.9P pv_pred (t), P pv_max (t) = 1.1P pv_pred (t);

[0037] The intra-day prediction model is used to output a corrected charge-discharge plan at a resolution of 15 minutes, based on a 24-hour SOC reference trajectory, a benchmark charge-discharge plan, and feedback data; the feedback data includes the actual SOC, photovoltaic output deviation, and frequency modulation response record;

[0038] The objective function output by the intra-day prediction model is to minimize the value of the prediction deviation function, and is constrained by the photovoltaic output constraint function and the power volatility limit function;

[0039] The formula of the prediction deviation function is:

[0040]

[0041] Among them, Min Cost ID represents the minimum value of the prediction deviation of the energy storage system; the value of k is 0, 1, 2...; λ ID (t) represents the predicted electricity price value at the t-th hour within a day; represents the discharge power of the energy storage system at the t-th hour within a day; α represents the SOC tracking weight coefficient, which is used to control the deviation penalty strength between the 24-hour SOC reference trajectory and the benchmark charge-discharge plan; SOC'(t) represents the SOC of the energy storage system at the t-th hour within a day; SOC' ref (t) represents the 24-hour SOC reference trajectory;

[0042] The formula of the photovoltaic output constraint function is:

[0043] P pv,ID (t) = P pv,DA (t) + ΔP pv,error (t);

[0044] Among them, P pv,ID (t) represents the photovoltaic output value at the t-th moment within a day; P pv,DA (t) represents the photovoltaic prediction value at the t-th moment of the day before; ΔP pv,error (t) represents the photovoltaic prediction error value at the t-th moment of the day before, and ΔP pv,error (t) ~ N(0, σ 2 ), σ represents the standard deviation of the normal distribution;

[0045] The formula of the power volatility limit function is:

[0046] |P' grid (t) - P' grid (t - 1)| ≤ 0.1P' grid,max ;

[0047] Among them, P' grid (t) represents the grid power at the t-th moment within a day; P' grid (t - 1) represents the grid power at the (t - 1)-th moment within a day; P' grid,max represents the maximum grid output power within a day;

[0048] The real-time prediction model is used to output charge-discharge instructions and frequency modulation output power based on the corrected charge-discharge plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output, and actual SOC at a resolution of 1 minute;

[0049] The objective function output by the real-time prediction model is to maximize the value of the electricity revenue function, and is constrained by the real-time power balance function and the frequency modulation power limit function;

[0050] The formula of the electricity revenue function is:

[0051]

[0052] Among them, Max Profitt RT represents the maximum value of electricity revenue; λ RT (t) represents the real-time electricity price; P” dischar (t) represents the discharge power of the energy storage system at time t; P” char (t) represents the charging power of the energy storage system at time t; Δt RT represents the control period of the real-time prediction model; β represents the frequency modulation revenue weight coefficient; P FR (t) represents the power responding to the grid frequency modulation demand at time t;

[0053] The formula of the real-time power balance function is as follows:

[0054] P” grid (t) = P” load (t) + P” char (t) - P pv,actual (t) - P” dischar (t) + P FR (t);

[0055] Among them, P” grid (t) represents the grid power at time t; P” load (t) represents the load power at time t; P” char (t) represents the charging power of the energy storage system at time t; P pv,actual (t) represents the actual PV output at time t;

[0056] The formula of the frequency modulation power limit function is as follows:

[0057] 0 ≤ P FR (t) ≤ 0.3P rated ;

[0058] Among them, P rated represents the rated power of the energy storage system.

[0059] Furthermore, the specific content of step S6 is as follows:

[0060] Based on a 24-hour sliding window, the output of the day-ahead prediction model is optimized by rolling at a frequency of once every 24 hours; based on a 4-hour sliding window, the output of the intraday prediction model is optimized by rolling at a frequency of once every 15 minutes; based on a 15-minute sliding window, the output of the real-time prediction model is optimized by rolling at a frequency of once every 1 minute.

[0061] Furthermore, in step S7, the deviation compensation mechanism is specifically as follows:

[0062] When the deviation of real-time photovoltaic output power is greater than 15%, or the volatility of real-time electricity price is greater than 20%, adjust the charging power and discharging power of the energy storage system based on a preset priority; based on the frequency modulation signal issued by the power grid, switch the energy storage system to the frequency modulation mode, and respond to the frequency deviation according to P FR (t) = K·Δf(t); where K represents the frequency modulation coefficient, and Δf(t) represents the power grid frequency offset.

[0063] In a second aspect, the present invention provides an energy storage regulation system for a photovoltaic energy storage based on multi-time scale rolling optimization, including the following modules:

[0064] A model creation module for creating and training an electricity price prediction model, a photovoltaic output prediction model, a load prediction model, a day-ahead prediction model, an intraday prediction model, and a real-time prediction model;

[0065] A data acquisition module for obtaining historical electricity price data, historical photovoltaic output data, historical load data, and the initial SOC of the energy storage system, inputting the historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, inputting the historical photovoltaic output data into the photovoltaic output prediction model to obtain a 24-hour photovoltaic output prediction curve, and inputting the historical load data into the load prediction model to obtain a 24-hour load prediction curve;

[0066] A day-ahead prediction module for inputting the 24-hour electricity price prediction curve, 24-hour photovoltaic output prediction curve, 24-hour load prediction curve, and the initial SOC into the day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a reference charge-discharge plan;

[0067] An intraday prediction module for collecting feedback data including the actual SOC, photovoltaic output deviation, and frequency modulation response records, inputting the 24-hour SOC reference trajectory, reference charge-discharge plan, and feedback data into the intraday prediction model to obtain a corrected charge-discharge plan;

[0068] A real-time prediction module for setting the SOC dynamic limit, collecting the real-time electricity price and real-time photovoltaic output, inputting the corrected charge-discharge plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output, and actual SOC into the real-time prediction model to obtain a charge-discharge instruction and a frequency modulation output power, and performing energy storage regulation based on the charge-discharge instruction and the frequency modulation output power;

[0069] A rolling optimization module for rollingly optimizing the outputs of the day-ahead prediction model, intraday prediction model, and real-time prediction model;

[0070] A deviation compensation module for setting a deviation compensation mechanism and compensating the frequency modulation output power based on the deviation compensation mechanism.

[0071] Further, in the model creation module, the electricity price prediction model is constructed based on the first LSTM network and is used to output a 24-hour electricity price prediction curve according to historical electricity price data;

[0072] The photovoltaic output prediction model is constructed based on the NWP meteorological network and the XGBoost network and is used to output a 24-hour photovoltaic output prediction curve according to historical photovoltaic output data;

[0073] The load prediction model is constructed based on the second LSTM network and is used to output a 24-hour load prediction curve according to historical load data.

[0074] Further, in the model creation module, the day-ahead prediction model is used to output a 24-hour SOC reference trajectory and a benchmark charge-discharge plan based on a 1-hour resolution, the 24-hour electricity price prediction curve, the 24-hour photovoltaic output prediction curve, the 24-hour load prediction curve, and the initial SOC; the 24-hour SOC reference trajectory serves as the boundary condition for the intra-day prediction model;

[0075] The objective function output by the day-ahead prediction model is to maximize the value of the full-cycle profit function, and is constrained by the energy storage dynamic constraint function, the grid interaction constraint function, the charge-discharge mutual exclusion constraint function, and the photovoltaic output uncertainty robust optimization constraint function;

[0076] The formula for the full-cycle profit function is:

[0077]

[0078] Among them, Max Profit DA represents the maximum value of the full-cycle profit of the energy storage system; λ DA (t) represents the predicted electricity price value at the t-th hour of the day-ahead; P dischar (t) represents the discharge power of the energy storage system at the t-th hour of the day-ahead; P char (t) represents the charging power of the energy storage system at the t-th hour of the day-ahead; Δt DA represents the time resolution of the day-ahead prediction model, that is, 1 hour; C degradation represents the life loss cost based on the Rainflow counting method; C grid represents the grid interaction penalty term, which is used to suppress the power peak-valley fluctuation;

[0079] The formula for the energy storage dynamic constraint function is:

[0080]

[0081] Among them, SOC(t) represents the SOC of the energy storage system at the t-th hour of the day before; SOC(t - 1) represents the SOC of the energy storage system at the (t - 1)-th hour of the day before; η char represents the charging efficiency of the energy storage system; represents the discharging efficiency of the energy storage system; E rated represents the rated capacity of the energy storage system;

[0082] The formula of the grid interaction constraint function is:

[0083] P grid,min (t) ≤ P grid (t) = P load (t) + P char (t) - P pv_pred (t) - P dischar (t) ≤

[0084] P grid,max (t);

[0085] Among them, P grid,min (t) represents the minimum grid output power at the t-th hour of the day before; P grid (t) represents the grid power at the t-th hour of the day before; P load (t) represents the load power at the t-th hour of the day before; P pv_pred (t) represents the predicted value of the photovoltaic output at the t-th hour of the day before; P grid,max (t) represents the maximum grid output power at the t-th hour of the day before;

[0086] The formula of the charge-discharge mutual exclusion constraint function is:

[0087] P dischar (t) · P char (t) = 0;

[0088] The formula of the photovoltaic output uncertainty robust optimization constraint function is:

[0089] P pv_min (t) ≤ P pv_pred (t) ≤ P pv_max (t);

[0090] Among them, P pv_min (t) represents the minimum value of the predicted value of the photovoltaic output at the t-th hour of the day before; P pv_max (t) represents the maximum value of the predicted value of the photovoltaic output at the t-th hour of the day before; and P pv_min (t) = 0.9P pv_pred (t), P pv_max (t) = 1.1P pv_pred (t);

[0091] The intra-day prediction model is used to output a corrected charge-discharge plan based on a 24-hour SOC reference trajectory, a benchmark charge-discharge plan, and feedback data at a resolution of 15 minutes; the feedback data includes the actual SOC, photovoltaic output deviation, and frequency regulation response record;

[0092] The objective function output by the intra-day prediction model is to minimize the value of the prediction deviation function, and is constrained by a photovoltaic output constraint function and a power volatility limit function;

[0093] The formula for the prediction deviation function is:

[0094]

[0095] where Min Cost ID represents the minimum value of the energy storage system prediction deviation; the value of k is 0, 1, 2...; λ ID (t) represents the predicted electricity price value at the t-th hour within the day; represents the discharge power of the energy storage system at the t-th hour within the day; α represents the SOC tracking weight coefficient, which is used to control the deviation penalty intensity between the 24-hour SOC reference trajectory and the benchmark charge-discharge plan; SOC'(t) represents the SOC of the energy storage system at the t-th hour within the day; SOC' ref (t) represents the 24-hour SOC reference trajectory;

[0096] The formula for the photovoltaic output constraint function is:

[0097] P pv,ID (t) = P pv,DA (t) + ΔP pv,error (t);

[0098] where P pv,ID (t) represents the photovoltaic output value at the t-th moment within the day; P pv,DA (t) represents the photovoltaic prediction value at the t-th moment before the day; ΔP pv,error (t) represents the photovoltaic prediction error value at the t-th moment before the day, and ΔP pv,error (t) ~ N(0, σ 2 ), σ represents the standard deviation of the normal distribution;

[0099] The formula for the power volatility limit function is:

[0100] |P' grid (t) - P' grid (t - 1)| ≤ 0.1P' grid,max ;

[0101] where P' grid (t) represents the grid power at the t-th moment within the day; P' grid(t - 1) represents the grid power at the (t - 1)th moment within a day; P' grid,max represents the maximum power output of the grid within a day;

[0102] The real - time prediction model is used to output charge - discharge instructions and frequency - modulation output power at a resolution of 1 minute, based on the corrected charge - discharge plan, SOC dynamic limit, real - time electricity price, real - time photovoltaic output, and actual SOC;

[0103] The objective function output by the real - time prediction model is to maximize the value of the electricity revenue function, and is constrained by the real - time power balance function and the frequency - modulation power limit function;

[0104] The formula for the electricity revenue function is:

[0105]

[0106] where, Max Profitt RT represents the maximum value of the electricity revenue; λ RT (t) represents the real - time electricity price; P” dischar (t) represents the discharge power of the energy storage system at the tth moment; P” char (t) represents the charge power of the energy storage system at the tth moment; Δt RT represents the control period of the real - time prediction model; β represents the frequency - modulation revenue weight coefficient; P FR (t) represents the power responding to the grid frequency - modulation demand at the tth moment;

[0107] The formula for the real - time power balance function is:

[0108] P” grid (t) = P” load (t)+P” char (t)-P pv,actual (t)-P” dischar (t)+P FR (t);

[0109] where, P” grid (t) represents the grid power at the tth moment; P” load (t) represents the load power at the tth moment; P” char (t) represents the charge power of the energy storage system at the tth moment; P pv,actual (t) represents the actual photovoltaic output at the tth moment;

[0110] The formula for the frequency - modulation power limit function is:

[0111] 0 ≤ P FR (t) ≤ 0.3P rated ;

[0112] where, Prated Represents the rated power of the energy storage system.

[0113] Furthermore, the rolling optimization module is specifically configured to:

[0114] Based on a 24-hour sliding window, optimize the output of the day-ahead prediction model at a frequency of once every 24 hours; based on a 4-hour sliding window, optimize the output of the intraday prediction model at a frequency of once every 15 minutes; based on a 15-minute sliding window, optimize the output of the real-time prediction model at a frequency of once every minute.

[0115] Furthermore, in the deviation compensation module, the deviation compensation mechanism is specifically:

[0116] When the real-time PV output deviation is greater than 15%, or the real-time electricity price volatility is greater than 20%, adjust the charging power and discharging power of the energy storage system based on a preset priority; based on the frequency modulation signal issued by the power grid, switch the energy storage system to the frequency modulation mode and respond to the frequency deviation according to P FR (t) = K·Δf(t); where K represents the frequency modulation coefficient, and Δf(t) represents the power grid frequency offset.

[0117] The advantages of the present invention are:

[0118] By creating and training an electricity price prediction model, a photovoltaic output prediction model, a load prediction model, a day-ahead prediction model, an intraday prediction model, and a real-time prediction model; then inputting historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, inputting historical photovoltaic output data into the photovoltaic output prediction model to obtain a 24-hour photovoltaic output prediction curve, and inputting historical load data into the load prediction model to obtain a 24-hour load prediction curve; inputting the 24-hour electricity price prediction curve, the 24-hour photovoltaic output prediction curve, the 24-hour load prediction curve, and the initial SOC into the day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a benchmark charge-discharge plan; then collecting feedback data including actual SOC, photovoltaic output deviation, and frequency regulation response records, and inputting the 24-hour SOC reference trajectory, the benchmark charge-discharge plan, and the feedback data into the intraday prediction model to obtain a corrected charge-discharge plan; then setting the SOC dynamic limit, collecting the real-time electricity price and the real-time photovoltaic output, and inputting the corrected charge-discharge plan, the SOC dynamic limit, the real-time electricity price, the real-time photovoltaic output, and the actual SOC into the real-time prediction model to obtain a charge-discharge command and a frequency regulation output power, and performing energy storage regulation based on the charge-discharge command and the frequency regulation output power; rolling and optimizing the outputs of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model, and compensating the frequency regulation output power based on the set deviation compensation mechanism; that is, performing energy storage regulation based on the multi-time scales (24 hours for day-ahead, 4 hours for intraday, and 15 minutes for real-time) of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model to overcome the problem of using a single time scale traditionally; dynamically adjusting the charging power and discharging power of the energy storage system through the deviation compensation mechanism to overcome the problem of cumulative prediction errors traditionally, and also avoiding frequent adjustment of the charge-discharge of the energy storage system and participating in frequency regulation to adapt to the power grid; by setting the constraint functions of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model, fully considering the battery life loss cost, avoiding sacrificing the economy of the entire life cycle of the energy storage system in pursuit of short-term benefits, and ultimately greatly improving the accuracy, economy, and power grid adaptability of energy storage regulation, and greatly delaying the battery life attenuation speed. Description of the Drawings

[0119] The present invention will be further described below with reference to the accompanying drawings in conjunction with embodiments.

[0120] Figure 1 It is a flowchart of a photovoltaic energy storage regulation method based on multi-time scale rolling optimization of the present invention.

[0121] Figure 2 It is a schematic structural diagram of a photovoltaic energy storage regulation system based on multi-time scale rolling optimization of the present invention. Detailed Embodiments

[0122] The overall idea of the technical solution in the embodiments of this application is as follows: Energy storage regulation is carried out based on multi-time scales (24 hours for day-ahead, 4 hours for intra-day, and 15 minutes for real-time) of day-ahead prediction models, intra-day prediction models, and real-time prediction models to overcome the problem of using a single time scale traditionally; a deviation compensation mechanism is used to dynamically adjust the charging power and discharging power of the energy storage system to overcome the problem of cumulative prediction errors traditionally and avoid frequent adjustment of the charging and discharging of the energy storage system, and participate in frequency modulation to adapt to the power grid; by setting constraint functions for the day-ahead prediction model, intra-day prediction model, and real-time prediction model, the cost of battery life loss is fully considered to avoid sacrificing the economy of the entire life cycle of the energy storage system in pursuit of short-term benefits, thereby improving the accuracy, economy, and power grid adaptability of energy storage regulation and delaying the battery life attenuation rate.

[0123] Please refer to Figures 1 to 2 as shown in the following, a preferred embodiment of a photovoltaic energy storage regulation method based on multi-time scale rolling optimization according to the present invention includes the following steps:

[0124] Step S1: The server creates and trains a electricity price prediction model, a photovoltaic power output prediction model, a load prediction model, a day-ahead prediction model, an intra-day prediction model, and a real-time prediction model; specifically, in implementation, the output priority of the real-time prediction model is higher than the outputs of the day-ahead prediction model and the intra-day prediction model to ensure emergency condition response;

[0125] Step S2: The server obtains historical electricity price data, historical photovoltaic power output data, historical load data, and the initial SOC of the energy storage system, inputs the historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, inputs the historical photovoltaic power output data into the photovoltaic power output prediction model to obtain a 24-hour photovoltaic power output prediction curve, and inputs the historical load data into the load prediction model to obtain a 24-hour load prediction curve;

[0126] Step S3: The server inputs the 24-hour electricity price prediction curve, 24-hour photovoltaic power output prediction curve, 24-hour load prediction curve, and initial SOC into the day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a benchmark charging and discharging plan;

[0127] Step S4: The server collects feedback data including actual SOC, photovoltaic power output deviation, and frequency modulation response records, and inputs the 24-hour SOC reference trajectory, benchmark charging and discharging plan, and feedback data into the intra-day prediction model to obtain a corrected charging and discharging plan;

[0128] Step S5: The server sets the dynamic SOC limit, collects the real-time electricity price and real-time photovoltaic output, inputs the corrected charge-discharge plan, dynamic SOC limit, real-time electricity price, real-time photovoltaic output, and actual SOC into the real-time prediction model to obtain the charge-discharge command and the frequency modulation output power, and performs energy storage regulation based on the charge-discharge command and the frequency modulation output power; generates a regulation log based on the charge-discharge command and the frequency modulation output power, encrypts the regulation log into an encrypted log, and stores and distributes the encrypted log for backup; the dynamic SOC limit means that the value of SOC falls within [SOC min , SOC max ;

[0129] The encryption process of the regulation log is specifically as follows: Obtain the current timestamp, calculate the hash value of the regulation log and the timestamp, encrypt the regulation log, timestamp, and hash value through the AES algorithm to obtain the first-level encrypted data, shift each character of the first-level encrypted data three positions to the right in a circular manner to obtain the second-level encrypted data, and encrypt the second-level encrypted data into an encrypted log through the RC6 algorithm; by combining the timestamp, hash value, AES algorithm, RC6 algorithm, displacement direction, and displacement number of bits, six-fold security measures are taken, greatly improving the security of the storage and backup of the regulation log, preventing it from being stolen and tampered with in plain text, and facilitating later traceability;

[0130] Step S6: The server rolls and optimizes the outputs of the day-ahead prediction model, intra-day prediction model, and real-time prediction model;

[0131] Step S7: The server sets a deviation compensation mechanism and compensates the frequency modulation output power based on the deviation compensation mechanism.

[0132] The present invention effectively improves the economic benefits of the energy storage system in the scenarios of real-time electricity price fluctuations and photovoltaic intermittency, reduces the impact of prediction errors on energy storage regulation, extends the battery life, and enhances the coordination ability between the energy storage system and the power grid through multi-time-scale rolling optimization, a constraint function for combined optimization of benefits considering battery life, and a deviation compensation mechanism (dynamic response mechanism).

[0133] In step S1, the electricity price prediction model is constructed based on the first LSTM network and is used to output a 24-hour electricity price prediction curve according to the historical electricity price data;

[0134] The photovoltaic output prediction model is constructed based on the NWP meteorological network and the XGBoost network and is used to output a 24-hour photovoltaic output prediction curve according to the historical photovoltaic output data;

[0135] The load prediction model is constructed based on the second LSTM network and is used to output a 24-hour load prediction curve according to the historical load data.

[0136] In the step S1, the day-ahead prediction model is used to output a 24-hour SOC reference trajectory and a benchmark charge-discharge plan based on a 24-hour electricity price prediction curve, a 24-hour photovoltaic output prediction curve, a 24-hour load prediction curve, and an initial SOC at a resolution of 1 hour; the 24-hour SOC reference trajectory serves as the boundary condition for the intra-day prediction model;

[0137] The objective function output by the day-ahead prediction model is to maximize the value of the full-cycle profit function, and is constrained by a energy storage dynamic constraint function, a grid interaction constraint function, a charge-discharge mutual exclusion constraint function, and a photovoltaic output uncertainty robust optimization constraint function;

[0138] The formula for the full-cycle profit function is:

[0139]

[0140] where, Max Profit DA represents the maximum value of the full-cycle profit of the energy storage system; λ DA (t) represents the predicted electricity price value at the t-th hour of the day-ahead; P dischar (t) represents the discharge power of the energy storage system at the t-th hour of the day-ahead; P char (t) represents the charge power of the energy storage system at the t-th hour of the day-ahead; Δt DA represents the time resolution of the day-ahead prediction model, i.e., 1 hour; C degradation represents the life loss cost based on the Rainflow counting method; C grid represents the grid interaction penalty term, which is used to suppress the power peak-valley fluctuation;

[0141]

[0142] C grid =Σ[μ peak ·max(P grid (t))+μ valley ·min(P grid (t))];

[0143] where, k represents the unit cycle cost of the battery; DoD t represents the charge-discharge depth at the t-th time; n represents the fitting coefficient, and the value range is (1.2, 1.5); μ peak and μ valley both represent the grid peak-valley power penalty coefficients;

[0144] The formula for the energy storage dynamic constraint function is:

[0145]

[0146] Among them, SOC(t) represents the SOC of the energy storage system at the t-th hour of the day before; SOC(t - 1) represents the SOC of the energy storage system at the (t - 1)-th hour of the day before; η char represents the charging efficiency of the energy storage system; represents the discharging efficiency of the energy storage system; E rated represents the rated capacity of the energy storage system;

[0147] The formula of the grid interaction constraint function is:

[0148] P grid,min (t) ≤ P grid (t) = P load (t) + P char (t) - P pv_pred (t) - P dischar (t) ≤

[0149] P grid,max (t);

[0150] Among them, P grid,min (t) represents the minimum grid output power at the t-th hour of the day before; P grid (t) represents the grid power at the t-th hour of the day before; P load (t) represents the load power at the t-th hour of the day before; P pv_pred (t) represents the predicted value of photovoltaic power output at the t-th hour of the day before; P grid,max (t) represents the maximum grid output power at the t-th hour of the day before;

[0151] The formula of the charge-discharge mutual exclusion constraint function is:

[0152] P dischar (t) · P char (t) = 0;

[0153] The formula of the photovoltaic power output uncertainty robust optimization constraint function is:

[0154] P pv_min (t) ≤ P pv_pred (t) ≤ P pv_max (t);

[0155] Among them, P pv_min (t) represents the minimum value of the predicted value of photovoltaic power output at the t-th hour of the day before; P pv_max (t) represents the maximum value of the predicted value of photovoltaic power output at the t-th hour of the day before; and P pv_min (t) = 0.9P pv_pred (t), P pv_max (t) = 1.1P pv_pred (t), that is, covering 90% of the prediction confidence interval;

[0156] The intraday prediction model is used to output a corrected charging and discharging plan based on a 24-hour SOC reference trajectory, a benchmark charging and discharging plan, and feedback data at a resolution of 15 minutes; the feedback data includes the actual SOC, the photovoltaic output deviation, and the frequency regulation response record; the corrected charging and discharging plan serves as the boundary condition for the real-time prediction model;

[0157] The objective function output by the intraday prediction model is to minimize the value of the prediction deviation function, and is constrained by the photovoltaic output constraint function and the power volatility limit function;

[0158] The formula for the prediction deviation function is:

[0159]

[0160] where Min Cost ID represents the minimum value of the prediction deviation of the energy storage system; the value of k is 0, 1, 2...; λ ID (t) represents the predicted electricity price value at the t-th hour within the day; represents the discharging power of the energy storage system at the t-th hour within the day; α represents the SOC tracking weight coefficient, which is used to control the deviation penalty strength between the 24-hour SOC reference trajectory and the benchmark charging and discharging plan; SOC'(t) represents the SOC of the energy storage system at the t-th hour within the day; SOC' ref (t) represents the 24-hour SOC reference trajectory;

[0161] The formula for the photovoltaic output constraint function is:

[0162] P pv,ID (t) = P pv,DA (t) + ΔP pv,error (t);

[0163] where P pv,ID (t) represents the photovoltaic output value at the t-th moment within the day; P pv,DA (t) represents the photovoltaic prediction value at the t-th moment before the day; ΔP pv,error (t) represents the photovoltaic prediction error value at the t-th moment before the day, and ΔP pv,error (t) ~ N(0, σ 2 ), σ represents the standard deviation of the normal distribution;

[0164] The formula for the power volatility limit function is:

[0165] |P' grid (t) - P' grid (t - 1)| ≤ 0.1P' grid,max ;

[0166] where P' grid (t) represents the grid power at the t-th moment within the day; P'grid (t - 1) represents the grid power at the (t - 1)-th moment within a day; P' grid,max represents the maximum power output of the grid within a day;

[0167] The real-time prediction model is used to output charge and discharge instructions and frequency modulation output power at a resolution of 1 minute based on the corrected charge and discharge plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output, and actual SOC;

[0168] The objective function output by the real-time prediction model is to maximize the value of the electricity revenue function, and is constrained by the real-time power balance function and the frequency modulation power limit function;

[0169] The formula for the electricity revenue function is:

[0170]

[0171] where, Max Profitt RT represents the maximum value of the electricity revenue; λ RT (t) represents the real-time electricity price; P” dischar (t) represents the discharge power of the energy storage system at the t-th moment; P” char (t) represents the charge power of the energy storage system at the t-th moment; Δt RT represents the control period of the real-time prediction model; β represents the frequency modulation revenue weight coefficient; P FR (t) represents the power responding to the grid frequency modulation demand at the t-th moment;

[0172] The formula for the real-time power balance function is:

[0173] P” grid (t) = P” load (t) + P” char (t) - P pv,actual (t) - P” dischar (t) + P FR (t);

[0174] where, P” grid (t) represents the grid power at the t-th moment; P” load (t) represents the load power at the t-th moment; P” char (t) represents the charge power of the energy storage system at the t-th moment; P pv,actual (t) represents the actual photovoltaic output at the t-th moment;

[0175] The formula for the frequency modulation power limit function is:

[0176] 0 ≤ P FR (t) ≤ 0.3P rated ;

[0177] Among them, P rated represents the rated power of the energy storage system.

[0178] The following further elaborates on the day-ahead prediction model, intra-day prediction model, and real-time prediction model of the present invention:

[0179] Day-ahead prediction model (24-hour rolling window): With a resolution of 1 hour, based on electricity price prediction and photovoltaic output prediction, a benchmark charge-discharge plan is generated.

[0180] Intra-day prediction model (4-hour rolling window): The prediction data is updated every 4 hours, and the sliding window mechanism is adopted to correct the benchmark charge-discharge plan to address short-term fluctuations in photovoltaic output and electricity price prediction deviation.

[0181] Real-time prediction model (15-minute rolling window): The real-time electricity price of the power grid (power market API interface) and the actual photovoltaic output are collected every 15 minutes, and the energy storage power is dynamically adjusted to compensate for prediction errors and respond to the power grid frequency regulation demand.

[0182] The specific step S6 is as follows:

[0183] Based on a 24-hour sliding window, the output of the day-ahead prediction model is optimized by rolling at a frequency of once every 24 hours; based on a 4-hour sliding window, the output of the intra-day prediction model is optimized by rolling at a frequency of once every 15 minutes; based on a 15-minute sliding window, the output of the real-time prediction model is optimized by rolling at a frequency of once every 1 minute.

[0184] That is, a rolling optimization closed-loop is constructed for each time level (day-ahead, intra-day, real-time), and the steps are as follows:

[0185] a. Prediction time domain division

[0186] Day-ahead: 24-hour sliding window, globally optimized once every 24 hours.

[0187] Intra-day: 4-hour sliding window, rolled once every 15 minutes, covering fine adjustment for the next 4 hours.

[0188] Real-time: 15-minute sliding window, rolled once every 1 minute, achieving second-level response.

[0189] b. Rolling execution steps

[0190] ① Initialization: Obtain the latest electricity price, photovoltaic output, and SOC status.

[0191] ② Rolling optimization: Call the prediction model to update the data in the future window, solve the objective function based on the current state, and generate a control sequence.

[0192] ③ Instruction execution: Only execute the power instruction at the first time, and use the predicted values for the remaining periods as a reference.

[0193] ④ Feedback update: Slide the window to the next moment and repeat steps ① - ③.

[0194] c. Time window parameters

[0195]

[0196] In the step S7, the deviation compensation mechanism is specifically:

[0197] When the real-time PV output deviation is greater than 15%, or the real-time electricity price volatility is greater than 20%, adjust the charging power and discharging power of the energy storage system based on the preset priority; based on the frequency modulation signal issued by the power grid, switch the energy storage system to the frequency modulation mode and respond to the frequency deviation according to P FR (t) = K·Δf(t); where K represents the frequency modulation coefficient, and Δf(t) represents the power grid frequency offset.

[0198] A preferred embodiment of an energy storage regulation system for a PV and energy storage system based on multi-time scale rolling optimization according to the present invention includes the following modules:

[0199] A model creation module, used for the server to create and train an electricity price prediction model, a PV output prediction model, a load prediction model, a day-ahead prediction model, an intraday prediction model, and a real-time prediction model; specifically in implementation, the output of the real-time prediction model has a higher priority than the outputs of the day-ahead prediction model and the intraday prediction model to ensure emergency condition response;

[0200] A data acquisition module, used for the server to obtain historical electricity price data, historical PV output data, historical load data, and the initial SOC of the energy storage system, input the historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, input the historical PV output data into the PV output prediction model to obtain a 24-hour PV output prediction curve, and input the historical load data into the load prediction model to obtain a 24-hour load prediction curve;

[0201] A day-ahead prediction module, used for the server to input the 24-hour electricity price prediction curve, 24-hour PV output prediction curve, 24-hour load prediction curve, and the initial SOC into the day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a reference charge and discharge plan;

[0202] An intraday prediction module, used for collecting feedback data including the actual SOC, PV output deviation, and frequency modulation response records, inputting the 24-hour SOC reference trajectory, reference charge and discharge plan, and feedback data into the intraday prediction model to obtain a corrected charge and discharge plan;

[0203] A real-time prediction module is used for the server to set the dynamic limit of SOC, collect the real-time electricity price and real-time photovoltaic output, input the corrected charge-discharge plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output and actual SOC into a real-time prediction model to obtain charge-discharge instructions and frequency modulation output power, and perform energy storage regulation based on the charge-discharge instructions and frequency modulation output power; generate a regulation log based on the charge-discharge instructions and frequency modulation output power, encrypt the regulation log into an encrypted log, and store and distribute the encrypted log for backup; the SOC dynamic limit means that the value of SOC falls within [SOC min , SOC max ;

[0204] The encryption process of the regulation log is specifically as follows: obtain the current timestamp, perform a hash calculation on the regulation log and the timestamp to obtain a hash value, encrypt the regulation log, timestamp and hash value through the AES algorithm to obtain first-level encrypted data, shift each character of the first-level encrypted data three positions to the right in a circular manner to obtain second-level encrypted data, and encrypt the second-level encrypted data into an encrypted log through the RC6 algorithm; by combining the timestamp, hash value, AES algorithm, RC6 algorithm, displacement direction, and displacement number of bits, six security measures are taken, greatly improving the security of the storage and backup of the regulation log, avoiding being stolen and tampered with in plain text, and facilitating later traceability;

[0205] A rolling optimization module is used for the server to roll and optimize the outputs of the day-ahead prediction model, intra-day prediction model and real-time prediction model;

[0206] A deviation compensation module is used for the server to set a deviation compensation mechanism and compensate the frequency modulation output power based on the deviation compensation mechanism.

[0207] The present invention effectively improves the economic benefits of the energy storage system in the scenarios of real-time electricity price fluctuations and photovoltaic intermittency, reduces the impact of prediction errors on energy storage regulation, extends the battery life, and simultaneously enhances the coordination ability between the energy storage system and the power grid through multi-time scale rolling optimization, a constraint function for combined optimization of benefits considering battery life, and a deviation compensation mechanism (dynamic response mechanism).

[0208] In the model creation module, the electricity price prediction model is constructed based on the first LSTM network and is used to output a 24-hour electricity price prediction curve according to historical electricity price data;

[0209] The photovoltaic output prediction model is constructed based on the NWP meteorological network and the XGBoost network and is used to output a 24-hour photovoltaic output prediction curve according to historical photovoltaic output data;

[0210] The load prediction model is constructed based on the second LSTM network and is used to output a 24-hour load prediction curve according to historical load data.

[0211] In the model creation module, the day-ahead prediction model is used to output a 24-hour SOC reference trajectory and a benchmark charge-discharge plan based on a 24-hour electricity price prediction curve, a 24-hour photovoltaic output prediction curve, a 24-hour load prediction curve, and an initial SOC at a resolution of 1 hour; the 24-hour SOC reference trajectory serves as the boundary condition for the intra-day prediction model.

[0212] The objective function output by the day-ahead prediction model is to maximize the value of the full-cycle profit function, which is constrained by an energy storage dynamic constraint function, a grid interaction constraint function, a charge-discharge mutual exclusion constraint function, and a photovoltaic output uncertainty robust optimization constraint function.

[0213] The formula for the full-cycle profit function is:

[0214]

[0215] where Max Profit DA represents the maximum value of the full-cycle profit of the energy storage system; λ DA (t) represents the predicted electricity price value at the t-th hour of the day-ahead; P dischar (t) represents the discharge power of the energy storage system at the t-th hour of the day-ahead; P char (t) represents the charging power of the energy storage system at the t-th hour of the day-ahead; Δt DA represents the time resolution of the day-ahead prediction model, i.e., 1 hour; C degradation represents the life loss cost based on the Rainflow counting method; C grid represents the grid interaction penalty term used to suppress power peak-valley fluctuations.

[0216]

[0217] C grid = ∑[μ peak · max(P grid (t)) + μ valley · min(P grid (t))];

[0218] where k represents the unit cycle cost of the battery; DoD t represents the charge-discharge depth at the t-th time; n represents the fitting coefficient with a value range of (1.2, 1.5); μ peak and μ valley both represent the grid peak-valley power penalty coefficients.

[0219] The formula for the energy storage dynamic constraint function is:

[0220]

[0221] Among them, SOC(t) represents the SOC of the energy storage system at the t-th hour of the day before; SOC(t - 1) represents the SOC of the energy storage system at the (t - 1)-th hour of the day before; η char represents the charging efficiency of the energy storage system; represents the discharging efficiency of the energy storage system; E rated represents the rated capacity of the energy storage system;

[0222] The formula of the grid interaction constraint function is:

[0223] P grid,min (t) ≤ P grid (t) = P load (t) + P char (t) - P pv_pred (t) - P dischar (t) ≤

[0224] P grid,max (t);

[0225] Among them, P grid,min (t) represents the minimum grid output power at the t-th hour of the day before; P grid (t) represents the grid power at the t-th hour of the day before; P load (t) represents the load power at the t-th hour of the day before; P pv_pred (t) represents the predicted value of the photovoltaic power output at the t-th hour of the day before; P grid,max (t) represents the maximum grid output power at the t-th hour of the day before;

[0226] The formula of the charge-discharge mutual exclusion constraint function is:

[0227] P dischar (t) · P char (t) = 0;

[0228] The formula of the photovoltaic power output uncertainty robust optimization constraint function is:

[0229] P pv_min (t) ≤ P pv_pred (t) ≤ P pv_max (t);

[0230] Among them, P pv_min (t) represents the minimum value of the predicted value of the photovoltaic power output at the t-th hour of the day before; P pv_max (t) represents the maximum value of the predicted value of the photovoltaic power output at the t-th hour of the day before; and P pv_min (t) = 0.9P pv_pred (t), P pv_max (t) = 1.1P pv_pred (t), that is, covering 90% of the prediction confidence interval;

[0231] The intraday prediction model is used to output a corrected charging and discharging plan based on a 24-hour SOC reference trajectory, a benchmark charging and discharging plan, and feedback data at a resolution of 15 minutes; the feedback data includes the actual SOC, photovoltaic output deviation, and frequency regulation response record; the corrected charging and discharging plan serves as the boundary condition for the real-time prediction model;

[0232] The objective function output by the intraday prediction model is to minimize the value of the prediction deviation function, and is constrained by the photovoltaic output constraint function and the power volatility limit function;

[0233] The formula for the prediction deviation function is:

[0234]

[0235] where Min Cost ID represents the minimum value of the prediction deviation of the energy storage system; the value of k is 0, 1, 2...; λ ID (t) represents the predicted electricity price value at the t-th hour within a day; represents the discharging power of the energy storage system at the t-th hour within a day; α represents the SOC tracking weight coefficient, which is used to control the deviation penalty intensity between the 24-hour SOC reference trajectory and the benchmark charging and discharging plan; SOC'(t) represents the SOC of the energy storage system at the t-th hour within a day; SOC' ref (t) represents the 24-hour SOC reference trajectory;

[0236] The formula for the photovoltaic output constraint function is:

[0237] P pv,ID (t) = P pv,DA (t) + ΔP pv,error (t);

[0238] where P pv,ID (t) represents the photovoltaic output value at the t-th moment within a day; P pv,DA (t) represents the photovoltaic prediction value at the t-th moment before the day; ΔP pv,error (t) represents the photovoltaic prediction error value at the t-th moment before the day, and ΔP pv,error (t) ~ N(0, σ 2 ), σ represents the standard deviation of the normal distribution;

[0239] The formula for the power volatility limit function is:

[0240] |P' grid (t) - P' grid (t - 1)| ≤ 0.1P' grid,max ;

[0241] where P' grid (t) represents the grid power at the t-th moment within a day; P'grid (t - 1) represents the grid power at the (t - 1)-th moment within a day; P' grid,max represents the maximum power output of the grid within a day;

[0242] The real-time prediction model is used to output charge-discharge instructions and frequency modulation output power based on the corrected charge-discharge plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output, and actual SOC at a resolution of 1 minute;

[0243] The objective function output by the real-time prediction model is to maximize the value of the electricity revenue function, and is constrained by the real-time power balance function and the frequency modulation power limit function;

[0244] The formula for the electricity revenue function is:

[0245]

[0246] where, Max Profitt RT represents the maximum value of the electricity revenue; λ RT (t) represents the real-time electricity price; P” dischar (t) represents the discharge power of the energy storage system at the t-th moment; P” char (t) represents the charging power of the energy storage system at the t-th moment; Δt RT represents the control period of the real-time prediction model; β represents the frequency modulation revenue weight coefficient; P FR (t) represents the power responding to the grid frequency modulation demand at the t-th moment;

[0247] The formula for the real-time power balance function is:

[0248] P” grid (t) = P” load (t) + P” char (t) - P pv,actual (t) - P” dischar (t) + P FR (t);

[0249] where, P” grid (t) represents the grid power at the t-th moment; P” load (t) represents the load power at the t-th moment; P” char (t) represents the charging power of the energy storage system at the t-th moment; P pv,actual (t) represents the actual photovoltaic output at the t-th moment;

[0250] The formula for the frequency modulation power limit function is:

[0251] 0 ≤ P FR (t) ≤ 0.3P rated ;

[0252] Among them, P rated represents the rated power of the energy storage system.

[0253] The day-ahead prediction model, intraday prediction model, and real-time prediction model of the present invention are further described below:

[0254] Day-ahead prediction model (24-hour rolling window): With a resolution of 1 hour, based on electricity price prediction and photovoltaic output prediction, a benchmark charge-discharge plan is generated.

[0255] Intraday prediction model (4-hour rolling window): The prediction data is updated every 4 hours, and the sliding window mechanism is used to correct the benchmark charge-discharge plan to solve the short-term fluctuations of photovoltaic output and the deviation of electricity price prediction.

[0256] Real-time prediction model (15-minute rolling window): The real-time electricity price of the power grid (power market API interface) and the actual photovoltaic output are collected every 15 minutes, and the energy storage power is dynamically adjusted to compensate for prediction errors and respond to the power grid frequency regulation requirements.

[0257] The rolling optimization module is specifically used for:

[0258] Based on a 24-hour sliding window, the output of the day-ahead prediction model is rolled and optimized at a frequency of once every 24 hours; based on a 4-hour sliding window, the output of the intraday prediction model is rolled and optimized at a frequency of once every 15 minutes; based on a 15-minute sliding window, the output of the real-time prediction model is rolled and optimized at a frequency of once every 1 minute.

[0259] That is, a rolling optimization closed loop is constructed for each time level (day-ahead, intraday, real-time), and the steps are as follows:

[0260] a. Prediction time domain division

[0261] Day-ahead: 24-hour sliding window, globally optimized once every 24 hours.

[0262] Intraday: 4-hour sliding window, rolled once every 15 minutes, covering fine adjustment of the next 4 hours.

[0263] Real-time: 15-minute sliding window, rolled once every 1 minute, achieving second-level response.

[0264] b. Rolling execution steps

[0265] ① Initialization: Obtain the latest electricity price, photovoltaic output, and SOC status.

[0266] ② Rolling optimization: Call the prediction model to update the data in the future window, solve the objective function based on the current state, and generate a control sequence.

[0267] ③ Instruction execution: Only execute the power instruction at the first time, and use the predicted values in the remaining periods as a reference.

[0268] ④ Feedback update: Slide the window to the next moment and repeat steps ① - ③.

[0269] c. Time window parameters

[0270]

[0271] In the deviation compensation module, the deviation compensation mechanism is specifically as follows:

[0272] When the real-time PV output deviation is greater than 15%, or the real-time electricity price volatility is greater than 20%, adjust the charging power and discharging power of the energy storage system based on the preset priority; based on the frequency modulation signal issued by the power grid, switch the energy storage system to the frequency modulation mode and respond to the frequency deviation according to P FR (t) = K·Δf(t); where K represents the frequency modulation coefficient, and Δf(t) represents the power grid frequency offset.

[0273] In summary, the advantages of the present invention are:

[0274] By creating and training an electricity price prediction model, a photovoltaic output prediction model, a load prediction model, a day-ahead prediction model, an intraday prediction model, and a real-time prediction model; then inputting historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, inputting historical photovoltaic output data into the photovoltaic output prediction model to obtain a 24-hour photovoltaic output prediction curve, and inputting historical load data into the load prediction model to obtain a 24-hour load prediction curve; inputting the 24-hour electricity price prediction curve, the 24-hour photovoltaic output prediction curve, the 24-hour load prediction curve, and the initial SOC into the day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a reference charge-discharge plan; then collecting feedback data including the actual SOC, photovoltaic output deviation, and frequency regulation response records, and inputting the 24-hour SOC reference trajectory, the reference charge-discharge plan, and the feedback data into the intraday prediction model to obtain a corrected charge-discharge plan; then setting the SOC dynamic limit, collecting the real-time electricity price and the real-time photovoltaic output, and inputting the corrected charge-discharge plan, the SOC dynamic limit, the real-time electricity price, the real-time photovoltaic output, and the actual SOC into the real-time prediction model to obtain a charge-discharge command and a frequency regulation output power, and performing energy storage regulation based on the charge-discharge command and the frequency regulation output power; rolling and optimizing the outputs of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model, and compensating the frequency regulation output power based on the set deviation compensation mechanism; that is, performing energy storage regulation based on the multi-time scales (24 hours for day-ahead, 4 hours for intraday, and 15 minutes for real-time) of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model to overcome the problem of using a single time scale traditionally; dynamically adjusting the charging power and discharging power of the energy storage system through the deviation compensation mechanism to overcome the problem of cumulative prediction errors traditionally, and also avoiding frequent adjustment of the charge and discharge of the energy storage system and participating in frequency regulation to adapt to the power grid; by setting the constraint functions of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model, fully considering the battery life loss cost, avoiding sacrificing the economy of the whole life cycle of the energy storage system for pursuing short-term benefits excessively, and ultimately greatly improving the accuracy, economy, and power grid adaptability of the energy storage regulation, and greatly delaying the battery life attenuation speed.

[0275] Although the specific embodiments of the present invention have been described above, those skilled in the art should understand that the specific embodiments we described are illustrative rather than used to limit the scope of the present invention. Equivalent modifications and variations made by those skilled in the art in accordance with the spirit of the present invention should be covered by the scope of the claims of the present invention.

Claims

1. A method for regulating and controlling photovoltaic energy storage based on multi-time scale rolling optimization, characterized in that: The steps include: Step S1, creating and training an electricity price prediction model, a photovoltaic output prediction model, a load prediction model, a day-ahead prediction model, a day-within-a-day prediction model, and a real-time prediction model; Step S2, obtaining electricity price historical data, photovoltaic output historical data, load historical data and the initial SOC of the energy storage system, inputting the electricity price historical data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, inputting the photovoltaic output historical data into the photovoltaic output prediction model to obtain a 24-hour photovoltaic output prediction curve, and inputting the load historical data into the load prediction model to obtain a 24-hour load prediction curve; Step S3, inputting the 24-hour electricity price forecast curve, the 24-hour photovoltaic output forecast curve, the 24-hour load forecast curve and the initial SOC into the day-ahead forecast model to obtain a 24-hour SOC reference trajectory and a benchmark charge and discharge plan; Step S4, collecting feedback data including actual SOC, photovoltaic output deviation and frequency modulation response record, inputting the 24-hour SOC reference trajectory, benchmark charge and discharge plan and feedback data into the intraday prediction model to obtain a revised charge and discharge plan; Step S5, setting the SOC dynamic limit, collecting the real-time electricity price and the real-time photovoltaic output, inputting the modified charge and discharge plan, the SOC dynamic limit, the real-time electricity price, the real-time photovoltaic output and the actual SOC into the real-time prediction model, obtaining the charge and discharge instructions and the frequency modulation output power, and performing energy storage control based on the charge and discharge instructions and the frequency modulation output power; Step S6, rolling optimization of the outputs of the day-ahead prediction model, the intraday prediction model, and the real-time prediction model; Step S7: setting a deviation compensation mechanism, and compensating the frequency modulation output power based on the deviation compensation mechanism.

2. The method for regulating and controlling photovoltaic energy storage based on multi-time scale rolling optimization according to claim 1, characterized in that: In the step S1, the electricity price prediction model is constructed based on the first LSTM network, and is used to output a 24-hour electricity price prediction curve based on electricity price historical data; The photovoltaic output prediction model is constructed based on the NWP meteorological network and the XGBoost network, and is used to output a 24-hour photovoltaic output prediction curve based on photovoltaic output historical data; The load forecasting model is constructed based on the second LSTM network and is used to output a 24-hour load forecasting curve based on historical load data.

3. The method for regulating and controlling photovoltaic energy storage based on multi-time scale rolling optimization according to claim 1, characterized in that: In step S1, the day-ahead prediction model is used to output a 24-hour SOC reference trajectory and a benchmark charge and discharge plan based on a 24-hour electricity price prediction curve, a 24-hour photovoltaic output prediction curve, a 24-hour load prediction curve and an initial SOC with a resolution of 1 hour; the 24-hour SOC reference trajectory is used as a boundary condition of the intraday prediction model; The objective function output by the day-ahead prediction model is to maximize the value of the full-cycle benefit function, and is constrained by the energy storage dynamic constraint function, the grid interaction constraint function, the charge-discharge mutual exclusion constraint function, and the photovoltaic output uncertainty robust optimization constraint function; The formula of the full-cycle profit function is: Among them, Max Profit DA Represents the maximum value of the full cycle benefit of the energy storage system; DA (t) represents the electricity price forecast value at the tth hour before the day; P dischar (t) represents the discharge power of the energy storage system at the tth hour before the day; P char (t) represents the charging power of the energy storage system at the tth hour before the day; Δt DA represents the time resolution of the day-ahead forecast model, i.e. 1 hour; C degradation represents the life loss cost based on the Rainflow counting method; C grid represents the grid interaction penalty term, which is used to suppress power peak and valley fluctuations; The formula of the energy storage dynamic constraint function is: Where SOC(t) represents the SOC of the energy storage system at the tth hour before the day; SOC(t-1) represents the SOC of the energy storage system at the t-1th hour before the day; η char Represents the charging efficiency of the energy storage system; η dischar Represents the discharge efficiency of the energy storage system; E rated Indicates the rated capacity of the energy storage system; The formula of the grid interaction constraint function is: P grid,min (t)≤P grid (t)=P load (t)+P char (t)-P pv_pred (t)-P dischar (t)≤ P grid,max (t); Among them, P grid,min (t) represents the minimum power output of the power grid at the t hour before the day; P grid (t) represents the power grid power at the tth hour before the day; P load (t) represents the load power at the tth hour before the day; P pv_pred (t) represents the predicted photovoltaic output value at the tth hour before the day; P grid,max (t) represents the maximum power output of the power grid at the t hour before the day; The formula of the charge-discharge mutual exclusion constraint function is: P dischar (t)·P char (t)=0; The formula of the photovoltaic output uncertainty robust optimization constraint function is: P pv_min (t)≤P pv_pred (t)≤P pv_max (t); Among them, P pv_min (t) represents the minimum value of the photovoltaic output forecast value at the t hour before the day; P pv_max (t) represents the maximum value of the predicted photovoltaic output at the t hour before the day; and P pv_min (t) = 0.9P pv_pred (t), P pv_max (t) = 1.1P pv_pred (t); The intraday prediction model is used to output a modified charge and discharge plan with a resolution of 15 minutes based on a 24-hour SOC reference trajectory, a benchmark charge and discharge plan, and feedback data; the feedback data includes actual SOC, photovoltaic output deviation, and frequency modulation response records; The objective function output by the intraday forecasting model is to minimize the value of the forecast deviation function, and is constrained by the photovoltaic output constraint function and the power fluctuation rate limit function; The formula of the prediction deviation function is: Among them, Min Cost ID Indicates the minimum value of the prediction deviation of the energy storage system; the value of k is 0, 1, 2...; λ ID (t) represents the predicted electricity price at hour t of the day; P' dischar (t) represents the discharge power of the energy storage system at the tth hour of the day; α represents the SOC tracking weight coefficient, which is used to control the penalty for deviation between the 24-hour SOC reference trajectory and the benchmark charge and discharge plan; SOC'(t) represents the SOC of the energy storage system at the tth hour of the day; SOC' ref (t) represents the 24-hour SOC reference trajectory; The formula of the photovoltaic output constraint function is: P pv,ID (t)=P pv,DA (t)+ΔP pv,error (t); Among them, P pv,ID (t) represents the photovoltaic output value at time t within the day; P pv,DA (t) represents the photovoltaic forecast value at time t the day before; ΔP pv,error (t) represents the photovoltaic prediction error value at time t a day before, and ΔP pv,error (t)~N(0,σ 2 ), σ represents the standard deviation of the normal distribution; The formula of the power fluctuation rate limiting function is: |P' grid (t)-P' grid (t-1)|≤0.1P' grid,max ; Among them, P' grid (t) represents the power grid power at time t within a day; P' grid (t-1) represents the power grid power at time t-1 within the day; P' grid,max Indicates the maximum power output of the power grid during the day; The real-time prediction model is used to output charging and discharging instructions and frequency-modulated output power with a resolution of 1 minute based on the modified charging and discharging plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output and actual SOC; The objective function output by the real-time prediction model is to maximize the value of the electricity fee revenue function, and is constrained by the real-time power balance function and the frequency modulation power limit function; The formula of the electricity fee income function is: Among them, Max Profitt RT Represents the maximum value of electricity revenue; RT (t) represents the real-time electricity price; P” dischar (t) represents the discharge power of the energy storage system at time t; P” char (t) represents the charging power of the energy storage system at time t; Δt RT represents the control period of the real-time prediction model; β represents the frequency modulation benefit weight coefficient; P FR (t) represents the power responding to the grid frequency regulation demand at time t; The formula of the real-time power balance function is: P” grid (t)=P” load (t)+P” char (t)-P pv,actual (t)-P” dischar (t)+P FR (t); Among them, P grid (t) represents the power of the power grid at time t; P” load (t) represents the load power at time t; P” char (t) represents the charging power of the energy storage system at time t; P pv,actual (t) represents the actual photovoltaic output at time t; The formula of the frequency modulation power limit function is: 0≤P FR (t)≤0.3P rated ; Among them, P rated Indicates the rated power of the energy storage system.

4. The method for regulating and controlling photovoltaic energy storage based on multi-time scale rolling optimization according to claim 1, characterized in that: The step S6 is specifically as follows: Based on a 24-hour sliding window, the output of the day-ahead prediction model is optimized once every 24 hours; based on a 4-hour sliding window, the output of the intraday prediction model is optimized once every 15 minutes; based on a 15-minute sliding window, the output of the real-time prediction model is optimized once every 1 minute.

5. The method for regulating and controlling photovoltaic energy storage based on multi-time scale rolling optimization according to claim 1, characterized in that: In step S7, the deviation compensation mechanism is specifically: When the real-time photovoltaic output deviation is greater than 15%, or the real-time electricity price fluctuation rate is greater than 20%, the charging power and discharging power of the energy storage system are adjusted based on the preset priority; based on the frequency modulation signal issued by the power grid, the energy storage system is switched to the frequency modulation mode and the P FR (t) = K·Δf(t) response frequency deviation; wherein K represents the frequency modulation coefficient, and Δf(t) represents the grid frequency offset.

6. A photovoltaic energy storage control system based on multi-time scale rolling optimization, characterized in that: Includes the following modules: A model creation module, used to create and train an electricity price prediction model, a photovoltaic output prediction model, a load prediction model, a day-ahead prediction model, a day-within-a-day prediction model, and a real-time prediction model; A data acquisition module is used to obtain historical electricity price data, historical photovoltaic output data, historical load data, and the initial SOC of the energy storage system, input the historical electricity price data into the electricity price prediction model to obtain a 24-hour electricity price prediction curve, input the historical photovoltaic output data into the photovoltaic output prediction model to obtain a 24-hour photovoltaic output prediction curve, and input the historical load data into the load prediction model to obtain a 24-hour load prediction curve; A day-ahead prediction module, for inputting the 24-hour electricity price prediction curve, the 24-hour photovoltaic output prediction curve, the 24-hour load prediction curve and the initial SOC into a day-ahead prediction model to obtain a 24-hour SOC reference trajectory and a benchmark charge and discharge plan; The intraday prediction module is used to collect feedback data including actual SOC, photovoltaic output deviation and frequency modulation response record, input the 24-hour SOC reference trajectory, benchmark charge and discharge plan and feedback data into the intraday prediction model to obtain a revised charge and discharge plan; A real-time prediction module is used to set the SOC dynamic limit, collect the real-time electricity price and the real-time photovoltaic output, input the modified charge and discharge plan, the SOC dynamic limit, the real-time electricity price, the real-time photovoltaic output and the actual SOC into the real-time prediction model, obtain the charge and discharge instructions and the frequency modulation output power, and perform energy storage control based on the charge and discharge instructions and the frequency modulation output power; A rolling optimization module, used for rolling optimization of the outputs of the day-ahead prediction model, the intraday prediction model and the real-time prediction model; The deviation compensation module is used to set a deviation compensation mechanism and compensate the frequency modulation output power based on the deviation compensation mechanism.

7. The photovoltaic energy storage control system based on multi-time scale rolling optimization according to claim 6, characterized in that: In the model creation module, the electricity price prediction model is constructed based on the first LSTM network, and is used to output a 24-hour electricity price prediction curve based on electricity price historical data; The photovoltaic output prediction model is constructed based on the NWP meteorological network and the XGBoost network, and is used to output a 24-hour photovoltaic output prediction curve based on photovoltaic output historical data; The load forecasting model is constructed based on the second LSTM network and is used to output a 24-hour load forecasting curve based on historical load data.

8. The photovoltaic energy storage control system based on multi-time scale rolling optimization according to claim 6, characterized in that: In the model creation module, the day-ahead prediction model is used to output a 24-hour SOC reference trajectory and a benchmark charge and discharge plan based on a 24-hour electricity price prediction curve, a 24-hour photovoltaic output prediction curve, a 24-hour load prediction curve, and an initial SOC with a resolution of 1 hour; the 24-hour SOC reference trajectory is used as a boundary condition for the intraday prediction model; The objective function output by the day-ahead prediction model is to maximize the value of the full-cycle benefit function, and is constrained by the energy storage dynamic constraint function, the grid interaction constraint function, the charge-discharge mutual exclusion constraint function, and the photovoltaic output uncertainty robust optimization constraint function; The formula of the full-cycle profit function is: Among them, Max Profit DA Represents the maximum value of the full cycle benefit of the energy storage system; DA (t) represents the predicted electricity price at the tth hour before the day; P dischar (t) represents the discharge power of the energy storage system at the tth hour before the day; P char (t) represents the charging power of the energy storage system at the tth hour before the day; Δt DA represents the time resolution of the day-ahead forecast model, i.e. 1 hour; C degradation represents the life loss cost based on the Rainflow counting method; C grid represents the grid interaction penalty term, which is used to suppress power peak and valley fluctuations; The formula of the energy storage dynamic constraint function is: Where SOC(t) represents the SOC of the energy storage system at the tth hour before the day; SOC(t-1) represents the SOC of the energy storage system at the t-1th hour before the day; η char It represents the charging efficiency of the energy storage system; it represents the discharging efficiency of the energy storage system; E rated Indicates the rated capacity of the energy storage system; The formula of the grid interaction constraint function is: P grid,min (t)≤P grid (t)=P load (t)+P char (t)-P pv_pred (t)-P dischar (t)≤ P grid,max (t); Among them, P grid,min (t) represents the minimum power output of the power grid at the t hour before the day; P grid (t) represents the power grid power at the tth hour before the day; P load (t) represents the load power at the tth hour before the day; P pv_pred (t) represents the predicted photovoltaic output value at the tth hour before the day; P grid,max (t) represents the maximum power output of the power grid at the t hour before the day; The formula of the charge-discharge mutual exclusion constraint function is: P dischar (t)·P char (t)=0; The formula of the photovoltaic output uncertainty robust optimization constraint function is: P pv_min (t)≤P pv_pred (t)≤P pv_max (t); Among them, P pv_min (t) represents the minimum value of the photovoltaic output forecast value at the t hour before the day; P pv_max (t) represents the maximum value of the predicted photovoltaic output at the t hour before the day; and P pv_min (t) = 0.9P pv_pred (t), P pv_max (t) = 1.1P pv_pred (t); The intraday prediction model is used to output a modified charge and discharge plan with a resolution of 15 minutes based on a 24-hour SOC reference trajectory, a benchmark charge and discharge plan, and feedback data; the feedback data includes actual SOC, photovoltaic output deviation, and frequency modulation response records; The objective function output by the intraday forecasting model is to minimize the value of the forecast deviation function, and is constrained by the photovoltaic output constraint function and the power fluctuation rate limit function; The formula of the prediction deviation function is: Among them, Min Cost ID Indicates the minimum value of the prediction deviation of the energy storage system; the value of k is 0, 1, 2...; λ ID (t) represents the electricity price forecast value at the tth hour of the day; represents the discharge power of the energy storage system at the tth hour of the day; α represents the SOC tracking weight coefficient, which is used to control the penalty for deviation from the 24-hour SOC reference trajectory and the benchmark charging and discharging plan; SOC'(t) represents the SOC of the energy storage system at the tth hour of the day; SOC' ref (t) represents the 24-hour SOC reference trajectory; The formula of the photovoltaic output constraint function is: P pv,ID (t)=P pv,DA (t)+ΔP pv,error (t); Among them, P pv,ID (t) represents the photovoltaic output value at time t within the day; P pv,DA (t) represents the photovoltaic forecast value at time t the day before; ΔP pv,error (t) represents the photovoltaic prediction error value at time t a day before, and ΔP pv,error (t)~N(0,σ 2 ), σ represents the standard deviation of the normal distribution; The formula of the power fluctuation rate limiting function is: |P' grid (t)-P' grid (t-1)|≤0.1P' grid,max ; Among them, P' grid (t) represents the power grid power at time t within a day; P' grid (t-1) represents the power grid power at time t-1 within the day; P' grid,max Indicates the maximum power output of the power grid during the day; The real-time prediction model is used to output charging and discharging instructions and frequency-modulated output power with a resolution of 1 minute based on the modified charging and discharging plan, SOC dynamic limit, real-time electricity price, real-time photovoltaic output and actual SOC; The objective function output by the real-time prediction model is to maximize the value of the electricity fee revenue function, and is constrained by the real-time power balance function and the frequency modulation power limit function; The formula of the electricity fee income function is: Among them, Max Profitt RT Represents the maximum value of electricity revenue; RT (t) represents the real-time electricity price; P” dischar (t) represents the discharge power of the energy storage system at time t; P” char (t) represents the charging power of the energy storage system at time t; Δt RT represents the control period of the real-time prediction model; β represents the frequency modulation benefit weight coefficient; P FR (t) represents the power responding to the grid frequency regulation demand at time t; The formula of the real-time power balance function is: P” grid (t)=P” load (t)+P” char (t)-P pv,actual (t)-P” dischar (t)+P FR (t); Among them, P grid (t) represents the power grid power at time t; P” load (t) represents the load power at time t; P” char (t) represents the charging power of the energy storage system at time t; P pv,actual (t) represents the actual photovoltaic output at time t; The formula of the frequency modulation power limit function is: 0≤P FR (t)≤0.3P rated ; Among them, P rated Indicates the rated power of the energy storage system.

9. The photovoltaic energy storage control system based on multi-time scale rolling optimization according to claim 6, characterized in that: The rolling optimization module is specifically used for: Based on a 24-hour sliding window, the output of the day-ahead prediction model is optimized once every 24 hours; based on a 4-hour sliding window, the output of the intraday prediction model is optimized once every 15 minutes; based on a 15-minute sliding window, the output of the real-time prediction model is optimized once every 1 minute.

10. The photovoltaic energy storage control system based on multi-time scale rolling optimization according to claim 6, characterized in that: In the deviation compensation module, the deviation compensation mechanism is specifically: When the real-time photovoltaic output deviation is greater than 15%, or the real-time electricity price fluctuation rate is greater than 20%, the charging power and discharging power of the energy storage system are adjusted based on the preset priority; based on the frequency modulation signal issued by the power grid, the energy storage system is switched to the frequency modulation mode and the P FR (t) = K·Δf(t) response frequency deviation; wherein K represents the frequency modulation coefficient, and Δf(t) represents the grid frequency offset.

Images