An optimization method and system suitable for source network load storage integrated base scheduling

Abstract

The application provides an optimization method and system suitable for source network load storage integrated base scheduling, the method comprises the following steps: S1, information updating and forecasting; S2, optimization model construction and collaborative strategy embedding; S3, optimization solving; S4, scheduling instruction issuing. The application focuses on the power optimization scheduling problem of the source network load storage integrated energy base, firstly, an optimization model with the minimum operation cost as the target is constructed, the core of which is to finely depict the mixed operation mechanism and dynamic constraints of the advanced adiabatic compressed air energy storage and the battery under the principle of'main-aid coordination'. Then, in order to effectively cope with the uncertainty of new energy output, the model predictive control framework is introduced to carry out rolling optimization and closed-loop correction on the scheduling plan, and the robustness of the strategy is improved. The strategy provided by the application provides an effective solution for realizing the economic and reliable operation of the source network load storage integrated base with high proportion of renewable energy.

Description

Technical Field

[0001] This invention belongs to the field of power system dispatching and automation technology, specifically relating to an optimization method and system suitable for integrated dispatching of power generation, grid, load and storage bases. Background Technology

[0002] In the process of building a new power system with new energy sources as the mainstay, the "integrated generation, grid, load, and storage" model is receiving widespread attention as a key path to achieve efficient local consumption of renewable energy. This integrated energy base aims to achieve regional power self-balancing through the coordinated dispatch of diverse resources such as wind power, photovoltaics, loads, and energy storage. However, truly achieving this goal faces dual challenges from both the system and equipment levels.

[0003] The primary challenge at the system level stems from the predictive uncertainty of renewable energy. The economic operation of an integrated facility heavily relies on a central dispatch system that formulates day-ahead or intraday optimized dispatch plans based on forecasts of future wind and solar output. However, the inherent errors in weather forecasts result in a persistent discrepancy between predicted and actual values. Traditional, deterministic "open-loop" dispatch methods (such as day-ahead optimization) generate fixed plans that are rigid and vulnerable to real-time forecast deviations, often leading to power imbalances, energy waste (wind and solar curtailment), and unexpected increases in operating costs, making it difficult to realize theoretically optimal economics in practical applications.

[0004] To address this uncertainty and fulfill the core task of balancing power, energy storage systems have become an indispensable component of integrated infrastructure. However, at the equipment level, the selection and application of energy storage technologies are inherently challenging. A single energy storage technology often cannot perfectly balance economy and flexibility; therefore, hybrid energy storage systems (HESS) composed of different technologies have emerged. One particularly promising combination is the integration of advanced adiabatic compressed air energy storage (AA-CAES) with battery energy storage (BESS).

[0005] The advantage of this combination lies in their theoretical perfect complementarity. AA-CAES, as a long-duration physical energy storage solution, boasts significant advantages in low cost and long lifespan, making it suitable for large-scale, hourly energy shifting and peak shaving. Battery energy storage, on the other hand, offers second-level rapid response capabilities, making it an ideal tool for handling high-frequency, transient power fluctuations. Combining the two theoretically allows the economic efficiency of AA-CAES to handle most of the energy throughput, while leveraging the flexibility of batteries to manage instantaneous fluctuations, thus constructing an economical and efficient energy storage solution.

[0006] However, translating this theoretical advantage into engineering practice presents significant technical challenges. The core difficulty lies in the vastly different and complex dynamic operating characteristics of the two technologies, which introduces immense complexity to coordinated scheduling: AA-CAES has a power response on the order of minutes, with minimum start-stop power thresholds and a slow power ramp-up rate, making it unable to independently complete rapid dynamic adjustment tasks. Furthermore, its energy conversion cycle efficiency is relatively low. Batteries, on the other hand, have extremely fast response times, but their investment costs are high and their cycle life is limited. Overusing batteries to handle energy time-shifting tasks that should be performed by AA-CAES will lead to premature aging, drastically increasing the overall system's operating and maintenance costs.

[0007] Therefore, the scheduling challenges of hybrid energy storage systems are intertwined with the aforementioned renewable energy forecasting bias problem, forming a key bottleneck that urgently needs to be overcome in the current technological field. An effective scheduling strategy must be able to solve problems at two levels simultaneously: first, at the micro level, it must perform real-time task allocation that "maximizes strengths and minimizes weaknesses" based on the dynamic characteristics and cost models of AA-CAES and batteries; second, at the macro level, it must ensure that this refined internal allocation strategy can dynamically adapt to real-time changes and errors in external wind and solar forecasts. Summary of the Invention

[0008] The first objective of this invention is to provide an optimized method for scheduling integrated power generation, grid, load and storage bases, addressing the aforementioned problems.

[0009] Therefore, the above-mentioned objective of the present invention is achieved through the following technical solution:

[0010] An optimization method for scheduling of integrated power generation, grid, load, and storage bases includes the following steps:

[0011] S1. Information Update and Prediction: Obtain the current status of each unit in the base, including the state of charge (SOC) of the battery and AA-CAES, and obtain the maximum available power of wind and solar power and load curves in the future prediction time domain.

[0012] S2. Optimization Model Construction and Cooperative Strategy Embedding: Establish a mathematical optimization model with the goal of minimizing total operating cost. The core of this model lies in finely characterizing the cooperative operation mechanism of the hybrid energy storage system (HESS), which consists of AA-CAES and batteries. The cooperative control strategy of "AA-CAES as the main component and batteries as the auxiliary component" is embedded in the model constraints.

[0013] S3. Optimization Solution: Solve the optimization problem in the current time domain and generate the optimal scheduling scheme covering the future time domain;

[0014] S4. Issuance of dispatch instructions: Only the first dispatch instruction in the plan is issued to the wind turbine, photovoltaic, hybrid energy storage and grid interaction unit for execution;

[0015] S5. Proceed to the next timing sequence to achieve closed-loop rolling optimization.

[0016] While adopting the above technical solutions, the present invention may also adopt or combine the following technical solutions:

[0017] As a preferred technical solution of the present invention: in step S2, the objective function includes grid power purchase cost, AA-CAES discharge cost, battery discharge cost, and hybrid energy storage charging incentive.

[0018] As a preferred technical solution of the present invention: in step S2, when constructing the optimization model, the following constraints must also be met: system power balance constraint, grid interaction power constraint, wind and solar power output constraint, and the respective state of charge (SOC) and charge / discharge power constraints of AA-CAES and battery energy storage.

[0019] As a preferred technical solution of the present invention: In step S2, when the total power command of the hybrid energy storage system is lower than the preset minimum start-up threshold of the AA-CAES system, the battery energy storage system independently undertakes the power command; when the total power command of the hybrid energy storage system is not lower than the preset minimum start-up threshold, the AA-CAES system undertakes the steady-state part of the power command, and the battery energy storage system compensates for the transient power difference caused by the dynamic characteristics limitation of the AA-CAES system.

[0020] As a preferred technical solution of the present invention, transient power difference compensation specifically includes: when the AA-CAES system performs positive power regulation, the battery energy storage system discharges to supplement the power gap; when the AA-CAES system performs negative power regulation, the battery energy storage system charges to absorb excess power.

[0021] The second objective of this invention is to provide an optimized system suitable for the scheduling of integrated power generation, grid, load, and storage bases, comprising the following modules:

[0022] The data acquisition and update module is used to obtain the current status of each unit in the base and to obtain the maximum available wind and solar power and load curves in the future prediction time domain.

[0023] The optimization model building module is used to establish a mathematical optimization model with the goal of minimizing the total operating cost.

[0024] The optimization and solution module is used to call optimization algorithms to solve the optimization problem in the current time domain and generate the optimal scheduling scheme covering the future time domain.

[0025] The dispatch instruction issuing module is used to issue the first dispatch instruction in the scheme to the wind turbine, photovoltaic, hybrid energy storage and grid interaction unit for execution;

[0026] The rolling optimization module is used to achieve closed-loop rolling optimization by repeating steps.

[0027] Compared with existing technologies, this invention has the following advantages: This invention focuses on the power optimization scheduling problem of integrated energy bases with power generation, grid, load, and storage. First, it constructs an optimization model with the goal of minimizing operating costs. Its core lies in finely characterizing the hybrid operation mechanism and dynamic constraints of advanced adiabatic compressed air energy storage and batteries under the principle of "main-auxiliary synergy." Subsequently, to effectively address the uncertainty of new energy output, a model predictive control framework is introduced to perform rolling optimization and closed-loop correction of the scheduling plan, improving the robustness of the strategy. The strategy proposed in this invention provides an effective solution for achieving the economical and reliable operation of integrated energy bases with a high proportion of renewable energy sources, grid, load, and storage. Attached Figure Description

[0028] Figure 1 This is a simplified topology diagram of the optimized system for integrated source-grid-load-storage base scheduling provided by the present invention.

[0029] Figure 2a The output power characteristic curve for AA-CAES positive regulation.

[0030] Figure 2b The output power characteristic curve for AA-CAES negative regulation.

[0031] Figure 3a This is a graph showing the power output of the HESS (Hybrid Energy Storage System) consisting of AA-CAES and battery cells over time when the hybrid energy storage output power is less than the minimum start-up power of AA-CAES.

[0032] Figure 3b This is a graph showing the power output curve of the HESS (Hybrid Energy Storage System) composed of AA-CAES and battery cells over time when the output power of the hybrid energy storage exceeds the lower limit of the starting power of AA-CAES.

[0033] Figure 4 The diagrams show the battery charge and discharge states under different AA-CAES output power variations. (a)-(c) represent positive AA-CAES power adjustment, and (d)-(f) represent negative AA-CAES power adjustment.

[0034] Figure 5 The flowchart for Model Predictive Control (MPC) is shown.

[0035] Figure 6 This is a graph showing the 24-hour predicted and actual maximum output curves of a photovoltaic power plant.

[0036] Figure 7 This is a graph showing the predicted and actual maximum output curves of a wind farm over 24 hours.

[0037] Figure 8 This is a 24-hour load curve.

[0038] Figure 9a This represents the real-time power output of each device cluster in Case 1.

[0039] Figure 9b The image shows the AA-CAES and lithium battery SOC state trajectory in Case 1.

[0040] Figure 10a This represents the real-time power output of each device cluster in Case 2.

[0041] Figure 10b This refers to the wind and solar power scheduling deviation in Case 2.

[0042] Figure 10c The image shows the AA-CAES and lithium battery SOC state trajectory in Case 2.

[0043] Figure 11a This represents the real-time power output of each device cluster in Case 3.

[0044] Figure 11b This is the SOC state trajectory of the lithium battery in Case 3.

[0045] Figure 12 This is a comparison chart of the total scheduling costs for Case 1, 2, and 3. Detailed Implementation

[0046] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0047] This method is applicable to integrated power generation, grid, load, and storage bases that have a weak connection with the power grid, such as... Figure 1 As shown, the hybrid energy storage system, loads, and new energy power plants are all connected to the same AC bus, and the arrows on the lines indicate the direction of power flow. To maintain power balance, a weak connection is established with the grid; that is, in the event of a power shortage, electricity is purchased from the grid, but surplus electricity is not sold back to the grid.

[0048] An optimization method for scheduling of integrated power generation, grid, load, and storage bases includes the following steps:

[0049] S1. Information Update and Prediction: Obtain the current status of each unit in the base, including the state of charge (SOC) of the battery and AA-CAES, and obtain the maximum available power of wind and solar power and load curves in the future prediction time domain.

[0050] The objective function includes grid purchase cost, AA-CAES discharge cost, battery discharge cost, and hybrid energy storage charging incentive.

[0051] S2. Optimization Model Construction and Cooperative Strategy Embedding: A mathematical optimization model is established with the goal of minimizing the total operating cost. The core of this model lies in its detailed characterization of the cooperative operation mechanism of the hybrid energy storage system (HESS). This hybrid energy storage system consists of AA-CAES and batteries. A cooperative control strategy of "AA-CAES as the main component and batteries as the auxiliary component" is embedded in the model constraints. This strategy determines whether the batteries respond independently or whether the AA-CAES dominates the steady-state power and the batteries provide transient assistance, based on the relationship between the total dispatch power command and the minimum start-up threshold of the AA-CAES.

[0052] When the total power command of the hybrid energy storage system is lower than the preset minimum start-up threshold of the AA-CAES system, the battery energy storage system independently undertakes the power command; when the total power command of the hybrid energy storage system is not lower than the preset minimum start-up threshold, the AA-CAES system undertakes the steady-state part of the power command, while the battery energy storage system compensates for the transient power difference caused by the dynamic characteristics limitation of the AA-CAES system.

[0053] The transient power difference compensation specifically includes: when the AA-CAES system performs positive power regulation, the battery energy storage system discharges to make up for the power gap; when the AA-CAES system performs negative power regulation, the battery energy storage system charges to absorb excess power.

[0054] When constructing the optimization model, the following constraints must also be met: system power balance constraints, grid interaction power constraints, new energy power station output constraints, and the respective state of charge (SOC) and charge / discharge power constraints of AA-CAES and battery energy storage.

[0055] S3. Optimization Solution: Solve the optimization problem in the current time domain and generate the optimal scheduling scheme covering the future time domain;

[0056] S4. Issuance of dispatch instructions: Only the first dispatch instruction in the scheme is issued to the wind turbine, photovoltaic, hybrid energy storage and grid interaction unit for execution, so as to realize closed-loop rolling optimization.

[0057] S5. Proceed to the next timing sequence to achieve closed-loop rolling optimization.

[0058] This invention also provides an optimized system for scheduling integrated power generation, grid, load, and storage bases, comprising the following modules:

[0059] The data acquisition and update module is used to obtain the current status of each unit in the base and to obtain the maximum available wind and solar power and load curves in the future prediction time domain.

[0060] The optimization model building module is used to establish a mathematical optimization model with the goal of minimizing the total operating cost.

[0061] The optimization and solution module is used to call optimization algorithms to solve the optimization problem in the current time domain and generate the optimal scheduling scheme covering the future time domain.

[0062] The dispatch instruction issuing module is used to issue the first dispatch instruction in the scheme to the wind turbine, photovoltaic, hybrid energy storage and grid interaction unit for execution;

[0063] The rolling optimization module is used to achieve closed-loop rolling optimization by repeating steps.

[0064] 1) Power scheduling framework:

[0065] The power dispatch framework of the integrated power generation, grid, load and storage base adopts centralized dispatch, that is, the central controller collects information from each unit, such as load power, hybrid energy storage SOC, and available power of new energy power stations. After the dispatch algorithm calculates, it issues power dispatch instructions, including hybrid energy storage charging and discharging power, new energy power station power generation power and power purchased from the grid.

[0066] The energy management principle of hybrid energy storage systems is "AA-CAES as the primary method, with battery energy storage as a supplement." Due to its low cost and long lifespan, AA-CAES can be used as the primary energy storage method, meeting all charge and discharge needs under steady-state conditions. However, when dealing with fluctuations in renewable energy sources, the minute-level, discontinuous dynamic response characteristics of AA-CAES are still insufficient. Therefore, the rapid response characteristics of battery energy storage need to be utilized as a supplement, playing an auxiliary role during transient processes. The output power characteristic curve of AA-CAES varies with operating conditions, such as... Figure 2a As shown, the positive regulation curve of AA-CAES shows that the change in output power is greater than or equal to zero; Figure 2b As shown, the negative regulation curve represents AA-CAES, with output power changes less than or equal to zero. The graph indicates that AA-CAES has a lower limit on its charging and discharging power, and the power conversion start-stop time between different operating conditions reaches the minute level, resulting in a piecewise linear charging and discharging power curve. In contrast, battery energy storage has no lower limit constraint within its regulation range and can achieve a power response in the second range. Therefore, to absorb the rapid power fluctuations of new energy sources, battery energy storage first responds quickly to supplement the power deficit, and then gradually withdraws as the power of AA-CAES increases, thus fully leveraging the advantages of both.

[0067] In fact, depending on the output power of the hybrid energy storage system, there are two coordination modes between AA-CAES and battery energy storage. The first mode is as follows: Figure 3a As shown, when the hybrid energy storage output power is less than the minimum starting power of AA-CAES, AA-CAES remains in standby mode, and the battery alone bears all the power output. The second mode is when the hybrid energy storage output power exceeds the lower limit of the AA-CAES starting power; in this case, AA-CAES is activated and operates accordingly, with the battery providing auxiliary output. For example... Figure 3bAs shown, the output power of AA-CAES changes relatively slowly due to limitations in start-up speed and ramp-up power. Under the aforementioned energy management principle of "AA-CAES as the main component and battery energy storage as a supplement," battery energy storage works in conjunction with AA-CAES to provide rapid response during transient processes; in steady state, the output power of battery energy storage is zero, and the output power of AA-CAES alone is equal to the output power of the hybrid energy storage system, thereby ensuring the system's rapid adjustment capability and efficient energy utilization.

[0068] 2) Power scheduling model:

[0069] The proposed optimization model consists of two parts: an objective function and a constraint set. The objective function uses the hybrid energy storage charging and discharging capacity and the grid-purchased electricity as variables, with the goal of minimizing cost. The constraint set defines the feasible domain of the variables and the operational limitations of the power plant equipment, ensuring the practical feasibility of the solution. Essentially, this optimization model seeks the optimal variable values ​​that maximize the objective function under given constraints.

[0070] 3) Objective function:

[0071] The objective function aims to minimize electricity costs. Since the cost of wind and solar power generation is near zero, their electricity prices are not considered in this model. Traditional electricity costs mainly consist of two parts: grid purchase costs and hybrid energy storage discharge costs. Grid purchase costs depend on the unit electricity price and electricity consumption at different times; hybrid energy storage discharge costs are determined by its discharge output, primarily including the discharge costs of AA-CAES and the battery.

[0072] Therefore, the objective function can be expressed as:

[0073] ;

[0074] In the formula, As a penalty factor; The total number of time scales; For grid electricity prices; for Electricity consumption of the power grid during a given time period; and These are AA-CAES and the levelized cost of electricity (LCOE) of the battery, respectively. and They are respectively The time period AA-CAES and the battery discharge level.

[0075] It should be noted that the above objective function may overlook the potential benefits of energy storage charging. For example, when wind and solar power generation is abundant, the system may forgo charging AA-CAES due to concerns about battery discharge costs, ultimately incurring higher expenditures due to electricity purchases during power shortages. Therefore, we incorporate energy storage charging as a positive incentive into the objective function:

[0076] ;

[0077] In the formula, This is the incentive coefficient; for Charging power of time-of-use hybrid energy storage system.

[0078] The formula for calculating the levelized cost of electricity (LCOE) of battery energy storage is as follows:

[0079] ;

[0080] In the formula, For initial capital expenditure; For the first Annual maintenance expenses; For the first Annual charging electricity price; and The first Annual charge / discharge volume; The discount rate; Project lifespan (years); Indicates the first to the second Summation of years; This is the discount factor.

[0081] 4) Constraints:

[0082] The optimal scheduling model must satisfy the following constraints to ensure the feasibility and practicality of the results:

[0083] (1) Power balance constraint

[0084] This constraint maintains a real-time balance between power generation and load:

[0085] ;

[0086] In the formula, for Power input to the power grid during a given period; and Wind turbines and photovoltaics respectively The amount of effort contributed during a given period; and For hybrid energy storage Discharge / charge power during the time period; for Load power during a given time period.

[0087] It should be noted that AA-CAES and the battery are in The charging and discharging power of different time periods should be considered uniformly, and the energy changes should be analyzed in conjunction with the hybrid energy storage control strategy.

[0088] (2) Power constraints of the power grid

[0089] The integrated power generation, grid, load, and storage base has a weak connection to the main power grid and follows two principles: 1) It does not sell electricity to the grid; 2) It only purchases electricity when its internal power generation is insufficient. Its power constraints are:

[0090] ;

[0091] In the formula, For time intervals.

[0092] (3) Power constraints of photovoltaic and wind turbines

[0093] The output of photovoltaic and wind turbines is affected by environmental factors, and their output range needs to be constrained to ensure system reliability.

[0094] ;

[0095] In the formula, and They are respectively Maximum available output of wind turbines and solar power during certain periods.

[0096] (4) Operational constraints of hybrid energy storage systems

[0097] Hybrid energy storage systems (HESS) are treated as a unified entity, with a focus on their external characteristics. Specific constraints on AA-CAES and batteries will be detailed in subsequent sections.

[0098] a) Charge and discharge power constraints

[0099] Under the energy management strategy of "AA-CAES as the primary and battery as the secondary," the HESS operating range is from zero to the maximum capacity of AA-CAES: when the scheduling command is lower than the minimum starting power of AA-CAES, only the battery responds; when it exceeds this threshold, AA-CAES handles steady-state demand, and the battery compensates for dynamic fluctuations. The constraints are as follows:

[0100] ;

[0101] In the formula, and It is a binary variable representing the charging and discharging state; and These represent the maximum power generation / compression power, respectively.

[0102] b) Operational status constraints

[0103] Hybrid energy storage systems cannot be charged and discharged simultaneously:

[0104] ;

[0105] (5) AA-CAES and battery operation constraints

[0106] The battery operates in two modes: independent operation or auxiliary AA-CAES. The former occurs when the dispatch command is lower than the minimum starting power of AA-CAES, and the battery responds to the demand independently; the latter is triggered when the command exceeds the threshold, and AA-CAES dominates the steady-state output, while the battery provides assistance by compensating for transient power deficits (positive regulation) or absorbing excess energy (reverse regulation).

[0107] a) Energy changes

[0108] i.AA-CAES is not activated.

[0109] At this time, the HESS output power is lower than the AA-CAES minimum starting power:

[0110] ;

[0111] In the formula, and These represent the minimum compression / generation power, respectively.

[0112] The battery independently processes commands, and its energy changes are as follows:

[0113] ;

[0114] In the formula, and They are respectively to Changes in battery charging / discharging energy over time.

[0115] AA-CAES energy remains unchanged:

[0116] ;

[0117] In the formula, They are respectively to Compressed / generated energy during a given period.

[0118] ii. AA-CAES forward / reverse adjustment

[0119] At this time, the HESS charging and discharging power exceeds the minimum starting power of AA-CAES:

[0120] ;

[0121] In steady state, the HESS output is dominated by AA-CAES:

[0122] ;

[0123] To simplify processing, the AA-CAES compression / generation power is uniformly expressed as:

[0124] ;

[0125] According to AA-CAES at time and The output power determines the power adjustment direction of the AA-CAES—positive or negative, such as... Figures 2a-2b As shown. Accordingly, the battery's operating states can be divided into two main categories: discharging and charging, encompassing a total of ten conditions, such as... Figure 4 As shown. For the first type, the AA-CAES receives a positive regulation command, resulting in an increase in output power ( During this period, the battery discharges to compensate for the transient power shortage, which can be divided into five different situations, such as... Figure 4 As shown in (a)-(c), the energy change expressions for the battery and AA-CAES are shown in Table 1.

[0126] Table 1. Conditions and expressions for battery discharge and AA-CAES energy change.

[0127]

[0128] Table 2. Battery charging and AA-CAES energy change conditions and expressions

[0129]

[0130] For the second type, the AA-CAES receives a negative regulation command, resulting in a decrease in output power. During this period, the battery needs to be charged to absorb excess transient power, which can be divided into five different situations, such as... Figure 4 As shown in (d)-(f). Time The expressions for battery charging energy and AA-CAES energy changes are listed in Table 2.

[0131] Here, ; , , , These are the compressor shutdown, generator start-up, generator shutdown, and compressor start-up times, respectively. , These represent the ramp-up rates for compression and power generation, respectively.

[0132] b) Energy constraint

[0133] In the second scenario, the current state of AA-CAES may be affected by the previous moment. To prevent overcharging / over-discharging, the battery energy change must satisfy:

[0134] ;

[0135] In the formula, and For the battery charge / discharge state binary variable (1 indicates activation); and for and Battery energy at all times; and For charge / discharge efficiency; and SOC limit; This refers to the battery capacity.

[0136] AA-CAES in The energy change over the period is as follows:

[0137] ;

[0138] In the formula, for AA-CAES energy at any moment; For compression / power generation efficiency; and SOC limit; For AA-CAES capacity.

[0139] c) Operational status constraints

[0140] The operating status of the battery and AA-CAES is handled in two categories.

[0141] i.AA-CAES is not activated.

[0142] When the energy storage operates independently, the battery state is consistent with that of HESS:

[0143] ;

[0144] AA-CAES is currently in a shutdown state.

[0145] ;

[0146] ii. AA-CAES forward / reverse adjustment

[0147] As an auxiliary unit, the state of the battery is determined by the power variation of the AA-CAES:

[0148] ;

[0149] AA-CAES status is synchronized with HESS:

[0150] ;

[0151] 5) Optimized scheduling strategy based on model predictive control:

[0152] Model predictive control (MPC) solves optimization problems repeatedly over a finite time domain by predicting the future behavior of a system. This method utilizes a mathematical model of the system, combined with current / historical data to predict the state, optimize the objective function, and satisfy system constraints. Its core is rolling optimization: only the first step of the control sequence is implemented, and iterative updates are made over time. This mechanism allows MPC to dynamically adapt to changing conditions, making it particularly suitable for real-time multi-objective optimization scenarios.

[0153] like Figure 5 As shown, the MPC scheduling strategy consists of six steps:

[0154] Step 1: Initialization and Information Update

[0155] In the control interval At the start time, the current state of the measurement system (Discrete state-space equations):

[0156] ;

[0157] in ;

[0158] ;

[0159] .

[0160] Simultaneously acquire the future Step Predicted Value

[0161] Step 2: Construct the prediction time-domain optimization problem

[0162] exist At any given moment, establish the prediction time domain. The optimization problem is to define each time step. System dynamics, constraints, and objective function:

[0163] ;

[0164] Step 3: Solve the finite-time optimization problem

[0165] Based on the current predictions, the optimal control sequence is obtained through optimization:

[0166] ;

[0167] Step 4: Implement initial control

[0168] Only the first step of the sequence is implemented. To accommodate future forecast biases.

[0169] Step 5: Closed-loop feedback and measurement

[0170] exist After the interval, measure Real-time system response:

[0171] ;

[0172] This feedback is used to correct the model's deviation from the prediction and improve the algorithm's robustness.

[0173] Step 6: Rolling Time Domain Update

[0174] As time advances, in Return to step 1 and re-execute at any time. By continuously integrating the latest forecast information, efficient and reliable real-time scheduling is achieved.

[0175] Case Study:

[0176] The superiority of the proposed scheduling algorithm is verified through economic comparison. First, the experimental configuration (including key parameters and wind and solar load data) is introduced, and then it is compared with two benchmark methods: 1) day-ahead optimization scheduling using hybrid energy storage (HABESS); 2) MPC scheduling using only battery energy storage.

[0177] 1) Data preparation

[0178] The installed capacity of photovoltaic / wind turbines is 200 MW / 150 MW, and the capacity of AA-CAES / lithium batteries is 120 MWh / 60 MWh. Line losses are ignored, and only power balance constraints are considered.

[0179] Table 3 Cost parameters of AA-CAES and lithium battery energy storage

[0180]

[0181] Table 4 Key parameters of the optimization model

[0182]

[0183] Key parameters are shown in Tables 3 and 4. AA-CAES and lithium battery LCOE are calculated using given formulas; SOC limits and efficiency are set based on engineering practice; the maximum power output curves for wind and solar power are generated using measured irradiance, temperature, and wind speed data from a specific day in 2023 over a 24-hour period, with predicted values ​​generated using a Gaussian error model; the load curves are derived from historical data from the same day, and due to their stable fluctuations (approximately 35 MW), the prediction error is negligible. The actual and predicted power output curves for wind and solar power are shown below. Figure 6 and Figure 7 As shown, the load fluctuation curve is as follows: Figure 8 As shown.

[0184] 2) Case Analysis:

[0185] Based on the same experimental foundation, this section compares three different operating scenarios:

[0186] Case 1 implements the proposed model predictive control (MPC)-based power optimal scheduling strategy, combining a hybrid advanced adiabatic compressed air energy storage and electrochemical energy storage system (HAB-ESS), and employs an optimized interval of 4 hours. Figure 9a As shown, the real-time power output of each device cluster (negative values ​​indicate the energy storage charging process) is as follows: Figure 9b As shown, AA-CAES corresponds to the state of charge (SOC) trajectory of a lithium battery.

[0187] like Figures 10a-10c As shown, Case 2 uses the same hybrid energy storage configuration, but employs a day-ahead globally optimal scheduling strategy based on predicted renewable energy generation. Figure 10b As shown, the wind power dispatch deviation caused by the discrepancy between forecasts and actual renewable energy output is clearly quantified (photovoltaics show no significant deviation).

[0188] Case 3 employs the same MPC strategy as Case 1, but uses only lithium-ion batteries for energy storage, maintaining a total energy storage capacity of 180 MWh. Its optimization parameters are identical to those of Case 1, and the relevant results are as follows: Figures 11a-11b As shown. Figure 12 A comparative summary of scheduling costs for three case studies is provided.

[0189] A comparative analysis of Case 1 and Case 2 demonstrates the superiority of the MPC-based real-time scheduling strategy in practical applications. Although their economic indicators are similar (136.7 k vs. 135.2 k), and day-ahead scheduling appears to have a slight cost advantage, the inherent prediction error makes it difficult to translate this theoretical advantage into practical benefits. Figure 10b As shown, when the actual output of renewable energy deviates from the predicted value, the day-ahead dispatch scheme will produce a continuous deviation in wind and solar power, which greatly reduces its theoretical optimality in actual operation.

[0190] In the comparison between Case 1 and Case 3, the economic advantages of the hybrid energy storage system are more prominent, with an operating cost reduction of approximately 18% (136.7 kWh vs. 161.5 kWh). This achievement mainly stems from synergistic optimization on two levels: First, by applying AA-CAES (386.43 yuan / MWh, 29.3% lower than lithium batteries) with lower discharge costs, it undertakes the baseload power supply task; second, the rapid response capability of lithium batteries (a ramp rate of 20% capacity / second) is utilized to handle short-term power fluctuations (duration less than 15 minutes). It is worth noting that although AA-CAES has inherent shortcomings such as lower efficiency (64% cycle efficiency, 26 percentage points lower than lithium batteries) and slower response (ramp rate less than 2% capacity / second), the system ultimately achieves stable power point tracking through the coordinated scheduling of the two energy storage technologies. This indicates that the improvement in the overall economic efficiency of the system does not rely on the performance breakthrough of a single device, but rather on the effective compensation of the limitations of AA-CAES by the dynamic adjustment capability of lithium batteries, forming a complementary operating pattern.

[0191] The case studies ultimately validated the superiority of the proposed method: compared with traditional day-ahead optimization, the model predictive control strategy can better adapt to actual operating conditions and resist prediction errors; compared with a single battery energy storage solution, the proposed hybrid energy storage scheduling strategy reduces the total daily scheduling cost by approximately 18% by leveraging the complementary advantages of different energy storage technologies.

[0192] The technical solution of the present invention has been described in conjunction with the specific experimental procedures shown in the accompanying drawings. However, the scope of protection of the present invention is not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions resulting from such changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An optimization method applicable to the scheduling of integrated power generation, grid, load, and storage bases, characterized in that, Includes the following steps: S1. Information Update and Prediction: Obtain the current status of each unit in the base, including the state of charge (SOC) of the battery and AA-CAES, and obtain the maximum available power of wind and solar power and load curves in the future prediction time domain. S2. Optimization Model Construction and Cooperative Strategy Embedding: A mathematical optimization model is established with the goal of minimizing total operating cost. This model describes the cooperative operation mechanism of the hybrid energy storage system (HESS), which consists of AA-CAES and batteries. A cooperative control strategy of "AA-CAES as the main component and batteries as the auxiliary component" is embedded in the model constraints. Depending on the output power of the hybrid energy storage system, there are two coordination modes between AA-CAES and battery energy storage. The first mode is when the output power of the hybrid energy storage is less than the minimum starting power of AA-CAES, AA-CAES remains in standby mode, and the battery alone undertakes all power output. The second mode is when the output power of the hybrid energy storage exceeds the lower limit of the starting power of AA-CAES. In this case, AA-CAES is activated and operates accordingly, and the battery provides auxiliary output. Incorporate the energy storage charging amount as a positive incentive into the objective function: ； In the formula, This is the incentive coefficient; for Charging power of time-of-use hybrid energy storage systems; As a penalty factor; The total number of time scales; For grid electricity prices; for Electricity consumption of the power grid during a given time period; and These are AA-CAES and the levelized cost of electricity (LCOE) of the battery, respectively. and They are respectively AA-CAES and battery discharge during the time period; S3. Optimization Solution: Solve the optimization problem in the current time domain and generate the optimal scheduling scheme covering the future time domain; Step 1: Initialization and Information Update In the control interval At the start time, the current state of the measurement system : ； in ; ； ； Simultaneously acquire the future Step Predicted Value ; Step 2: Construct the prediction time-domain optimization problem exist At any given moment, establish the prediction time domain. The optimization problem is to define each time step. The system dynamics, constraints, and objective function; Step 3: Solve the finite-time optimization problem Based on the current predictions, the optimal control sequence is obtained through optimization: ； S4. Issuance of dispatch instructions: Only the first dispatch instruction in the plan is issued to the wind turbine, photovoltaic, hybrid energy storage and grid interaction unit for execution; S5. Proceed to the next timing sequence to achieve closed-loop rolling optimization.

2. The method according to claim 1, characterized in that: In step S2, the objective function includes grid purchase cost, AA-CAES discharge cost, battery discharge cost, and hybrid energy storage charging incentive.

3. The method according to claim 1, characterized in that: In step S2, when constructing the optimization model, the following constraints must also be satisfied: system power balance constraints, grid interaction power constraints, wind and solar power output constraints, and the respective state of charge (SOC) and charge / discharge power constraints of AA-CAES and battery energy storage, specifically: 1) Power balance constraints This constraint maintains a real-time balance between power generation and load: ； In the formula, for Power input to the power grid during a given period; and Wind turbines and photovoltaics respectively The amount of effort contributed during a given period; and For hybrid energy storage Discharge / charge power during the time period; for Time-of-use load power; 2) Power constraints of the power grid The integrated power generation, grid, load, and storage base has a weak connection with the main power grid and follows two principles: it does not sell electricity to the grid; it only purchases electricity when its internal power generation is insufficient, and its power constraint is: ； In the formula, For time intervals; for Electricity consumption of the power grid during a given time period; 3) Power constraints of photovoltaic and wind turbines The output of photovoltaic and wind turbines is affected by environmental factors, and their output range needs to be constrained to ensure system reliability. ； In the formula, and They are respectively Maximum available output of wind turbines and solar power during certain periods; 4) Operational constraints of hybrid energy storage systems The hybrid energy storage system (HESS) is considered as a unified entity, with a focus on its external characteristics. a) Charge and discharge power constraints Under the energy management strategy of "AA-CAES as the primary and battery as the secondary", the HESS operating range is from zero to the maximum capacity of AA-CAES: when the scheduling command is lower than the minimum starting power of AA-CAES, only the battery responds; when it exceeds this threshold, AA-CAES handles steady-state demand, and the battery compensates for dynamic fluctuations, with the following constraints: ； In the formula, and It is a binary variable representing the charging and discharging state; and These are the maximum power generation / compression power, respectively; b) Operational status constraints Hybrid energy storage systems cannot be charged and discharged simultaneously: 。 4. The method according to claim 1, characterized in that: In step S2, when the total power command of the hybrid energy storage system is lower than the preset minimum start-up threshold of the AA-CAES system, the battery energy storage system independently undertakes the power command; when the total power command of the hybrid energy storage system is not lower than the preset minimum start-up threshold, the AA-CAES system undertakes the steady-state part of the power command, and the battery energy storage system compensates for the transient power difference caused by the dynamic characteristics limitation of the AA-CAES system.

5. The method according to claim 4, characterized in that: The transient power difference compensation specifically includes: when the AA-CAES system performs positive power regulation, the battery energy storage system discharges to make up for the power gap; when the AA-CAES system performs negative power regulation, the battery energy storage system charges to absorb excess power.

6. An optimized system suitable for the scheduling of integrated power generation, grid, load, and storage bases, characterized in that, The optimization system is based on the method as described in claim 1, and includes the following modules: The data acquisition and update module is used to obtain the current status of each unit in the base and to obtain the maximum available wind and solar power and load curves in the future prediction time domain. The optimization model building module is used to establish a mathematical optimization model with the goal of minimizing the total operating cost. The optimization and solution module is used to call optimization algorithms to solve the optimization problem in the current time domain and generate the optimal scheduling scheme covering the future time domain. The dispatch instruction issuing module is used to issue the first dispatch instruction in the scheme to the wind turbine, photovoltaic, hybrid energy storage and grid interaction unit for execution; The rolling optimization module is used to achieve closed-loop rolling optimization by repeating steps.

Images