Compressor energy storage peak shaving path adaptive planning and emergency capacity reservation method and system for extreme climate scenarios

Abstract

This invention relates to the field of novel power system operation and control technology, and discloses a method and system for adaptive planning of compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios. The method includes: collecting state data to construct feature vectors; calculating a conservative upper bound for net load by outputting load, renewable energy prediction values, and prediction uncertainties through a joint prediction model; constructing an extreme climate risk index; calculating available regulation capacity based on gas storage pressure, thermal storage status, ramp-up capability, and continuous support time, and dynamically dividing the capacity; and constructing a rolling robust optimization model to generate charging and discharging paths and reserve plans, and making rolling corrections. This invention drives dynamic partitioning and rolling correction of compressed gas storage capacity through the extreme climate risk index, which can improve the adaptability of peak-shaving paths under extreme climate conditions, avoid excessive consumption of emergency resources by conventional peak-shaving, and improve the operational resilience and renewable energy absorption capacity of novel power systems.

Description

Technical Field

[0001] This invention relates to the fields of new power system operation control, large-scale energy storage dispatch and grid security support technology, specifically to a method and system for adaptive planning of compressed gas energy storage peak shaving paths and emergency capacity reservation for extreme climate scenarios. Background Technology

[0002] With the continuous advancement of new power system construction, the installed capacity of new energy sources such as wind power and photovoltaics is growing rapidly, leading to greater volatility, randomness, and extremes in the grid's net load curve. On the one hand, periods of high renewable energy generation can easily result in low net load and pressure from wind and solar curtailment. On the other hand, under extreme weather conditions such as high temperatures, cold waves, continuous rain, typhoons, sandstorms, and icing, load demand may surge rapidly while renewable energy output may decline simultaneously, leading to risks such as insufficient peak-shaving resources, tight reserve capacity, and localized supply-demand imbalances in the system.

[0003] Compressed gas storage (CGS) is characterized by large capacity, long duration, long lifespan, and good safety, making it suitable for long-term regulation, deep peak shaving, backup support, and emergency response. Large-scale CGS power plants can absorb compressed air to generate electricity during periods of renewable energy surplus and release high-pressure air to generate electricity when the system is under high load or renewable energy is insufficient, thereby smoothing the net load and improving renewable energy absorption capacity. However, CGS is not an ideal battery model; its usability is affected by multiple factors, including gas pressure, thermal storage conditions, compressor and expander operating conditions, start-stop constraints, ramp rate, and boundary conditions.

[0004] Existing peak-shaving operation methods mostly rely on day-ahead load forecasts and renewable energy forecasts to formulate fixed charge-discharge plans, or use simple state-of-charge models to describe the available energy storage capacity. These methods are feasible to some extent under normal weather and typical daytime scenarios, but they have the following shortcomings under extreme weather conditions: First, the coupling relationship between weather, load and new energy output is not fully utilized, making it difficult to identify load gaps and low new energy output risks under extreme weather conditions such as cold waves, high temperatures, and continuous rain, resulting in a lack of foresight in dispatching plans.

[0005] Second, peak-shaving plans and emergency capacity reserves are usually formulated separately, lacking a mechanism for dynamic trade-offs within the same optimization framework. When the risk of extreme weather increases, energy storage capacity may be overused for routine peak shaving, resulting in insufficient emergency reserves during extreme events and severely impacting the reliability of the power supply system.

[0006] Third, the physical boundaries of compressed gas energy storage, such as gas pressure, thermal state, ramp rate, and continuous support time, are not fully incorporated into capacity reservation decisions. Existing methods often simply equate compressed gas energy storage to battery energy storage models, ignoring the unique physical characteristics of compressed gas energy storage, such as gas pressure leakage and thermal decay, which makes the strategy infeasible or unsafe in actual power plants.

[0007] Fourth, scheduling plans are mostly executed within fixed time periods, lacking adaptive correction based on rolling forecast errors. When actual operational deviations are large or risks change rapidly, fixed plans cannot be adjusted in a timely manner, resulting in deterioration of peak-shaving effectiveness or insufficient emergency capacity.

[0008] Fifth, the coordinated regulation relationship between compressed gas energy storage and conventional power sources such as thermal power and hydropower is not clear enough, making it difficult to form a regional integrated regulation mode, which limits the supporting potential of compressed gas energy storage in the large power grid.

[0009] Therefore, it is necessary to propose a peak-shaving path planning method that can simultaneously consider meteorological conditions, load, new energy output, and the physical constraints of compressed gas storage, in order to solve the technical problems in the existing technology of disconnect between peak-shaving plans and emergency reserves under extreme weather conditions, neglect of the physical boundaries of compressed gas storage, and lack of rolling adaptive correction of scheduling strategies. Summary of the Invention

[0010] To address the shortcomings of existing technologies, this invention provides a method and system for adaptive planning of compressed gas storage peak-shaving paths and emergency capacity reservation for extreme weather scenarios. This method can dynamically adjust the charging and discharging paths of compressed gas storage and the reserve capacity ratio based on extreme weather risks, ensuring peak-shaving effectiveness and renewable energy consumption while maintaining sufficient emergency support capabilities.

[0011] This invention is achieved through the following technical solution: An adaptive planning method for peak-shaving pathways and emergency capacity reservation for compressed gas storage in extreme climate scenarios is provided, including the following steps: Collect meteorological forecast data, historical load data, renewable energy output data, and compressed gas storage operation status data for the target time period, and construct the feature vector z for time period t. t eigenvector z t It should include at least meteorological elements, historical loads, historical renewable energy output, and calendar characteristics; Through the joint prediction model F θ Output load forecast values ​​for each time period within the future rolling window. Wind power forecast Photovoltaic forecast values and forecast uncertainty And calculate the upper limit of the conservative net load. , Confidence coefficient; Constructing an extreme climate risk index : ; Among them, T t W t G tThese are temperature, wind speed, and irradiance, respectively. t To predict the supply-demand gap, a1~a5 are weighting coefficients, and sat(·) is the limiting function; Gas storage pressure p based on compressed gas energy storage t Thermal storage status, climbing ability, and continuous support time; calculation of available adjustment capacity for time period t. and will Dynamically divided into peak-shaving capacity Frequency modulation reserve capacity Extreme weather emergency capacity and safety minimum capacity ,in: ; α0~α2 are the distribution coefficients. Based on the baseline net load, This is the rated power of compressed air energy storage; A rolling robust optimization model is constructed with the objectives of smoothing net load, reducing renewable energy curtailment, ensuring sufficient emergency capacity, and minimizing operating costs to generate the charging power of compressed gas storage. Discharge power and a backup plan, and meet the backup constraint; ; in, As a backup for conventional units, Reserved for demand response; Backup required for the system under extreme weather conditions: ; β1~β3 are reserve coefficients; Let the actual net load be N. t real Define prediction bias The equivalent energy state of compressed gas energy storage is denoted as E. t The gas storage pressure is denoted as p. t ;ε N ε ρ These are the net load deviation threshold and the risk change threshold, respectively. safe p safe These are the preset safe energy margin and safe gas storage pressure threshold, respectively; Forecast data and extreme climate risk index are updated on a rolling basis. And compressed gas energy storage state, when satisfied Time-triggered re-optimization; Output peak shaving routes, emergency capacity reservation plans, and risk warning levels.

[0012] Furthermore, the eigenvector zt Including temperature T t Wind speed W t Irradiance G t Humidity H t Precipitation or disaster warning intensity R t Historical load Historical wind power output Historical photovoltaic power output and features of date, time period and holidays. .

[0013] Furthermore, the equation of state for compressed gas energy storage includes the equivalent energy state E. t and gas storage pressure p t : ; ; in and These represent charging and discharging efficiencies, respectively. , , This is the pressure dynamic coefficient. Due to environmental pressures, This refers to gas leakage from the gas storage chamber. The operational constraints for compressed gas energy storage are: ; ; ; in , These are the maximum charging power and the maximum discharging power, respectively. , These are the upper and lower limits of the equivalent energy, respectively. , These are the upper and lower limits of the gas storage pressure. , These represent the charging and discharging ramp rates, respectively.

[0014] Furthermore, the objective function for rolling robust optimization is: ; in For real-time net load, Based on net load, For the abandoned electricity of new energy sources, As a penalty for insufficient emergency capacity, Δ The change in compressed gas energy storage power is represented by w1~w5, which are the weights.

[0015] Furthermore, when the risk of extreme weather increases... Increase or conservatively limit the upper limit of net load Exceeding the baseline net load At the same time, extreme weather emergency capacity Automatically increase and peak shaving capacity The corresponding compression will reduce the gas storage pressure p before the arrival of the high-risk period. t and safety minimum capacity Maintain a margin; release gradually once the risk is eliminated. Resume regular peak shaving.

[0016] Furthermore, the joint prediction model F θ The model can be one of the following: random forest, gradient boosting tree, long short-term memory network, Transformer, or mechanistic data fusion model. The rolling robust optimization adopts a model prediction control framework with a rolling window of 24h or 48h and a control period of 1h or 15min.

[0017] This invention also provides a system using an adaptive planning method for compressed gas storage peak-shaving paths and an emergency capacity reservation method for extreme climate scenarios, comprising: The data acquisition module is used to collect data on weather forecasts, load, new energy output, compressed gas storage status, and grid constraints. The joint forecasting module is used to perform joint forecasting and output load forecast values. Wind power forecast Photovoltaic forecast values and forecast uncertainty ; Extreme climate risk identification module, used to perform ρ t calculate; The compressed gas energy storage status assessment module is used to obtain gas storage pressure, thermal storage status, available power, ramp-up capability, and continuous support time. The capacity dynamic partitioning module is used to perform... Four-zone division and calculate; The peak shaving path optimization module is used to perform rolling robust optimization and output charging power. Discharge power and backup plans; The multi-power source collaborative verification module is used to combine the regulation power of thermal power, hydropower, demand response and electrochemical energy storage, and to perform regional comprehensive regulation and verification of the peak-shaving path and reserve plan of compressed gas energy storage. The rolling correction module is used to perform triggered re-optimization. The operation strategy output module is used to output peak shaving paths, emergency capacity reservation plans, early warning levels, and strategy explanations.

[0018] Furthermore, the capacity dynamic zoning module, based on the ρ output by the extreme climate risk identification module, t and the output of the joint prediction module Calculate emergency capacity for extreme weather The calculation results are then output to the peak shaving path optimization module; the triggering conditions for the rolling correction module include... ,in and The minimum available pressure of the compressed gas storage chamber and the energy required for emergency support are pre-calibrated and stored in the compressed gas storage status assessment module.

[0019] The present invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for adaptive planning of compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios.

[0020] The beneficial effects of this invention are: I. This invention constructs an extreme climate risk index, which uniformly quantifies meteorological deviation (temperature, wind speed, irradiance), forecast uncertainty, and supply-demand gap into risk indicators in the range of [0, 1], and explicitly embeds them into the emergency capacity calculation formula. When the extreme climate risk increases, the emergency capacity automatically increases and the peak-shaving capacity is correspondingly compressed, so that compressed gas storage can proactively retain gas and heat storage margins before the arrival of high-risk periods, avoiding excessive consumption of emergency resources by conventional peak shaving. This solves the problem of the disconnect between peak shaving plans and emergency reserves in the existing technology, and realizes adaptive emergency capacity reservation driven by extreme climate risk, avoiding excessive consumption of emergency resources by conventional peak shaving.

[0021] Second, this invention introduces a gas storage pressure state equation with a gas leakage term, and incorporates gas storage pressure, thermal storage state, ramp-up capability, and continuous support time into capacity reservation decisions and rolling correction criteria. Unlike traditional methods that simply equate compressed gas energy storage to battery energy storage, this invention accurately characterizes the unique physical characteristics of compressed gas energy storage, such as slow pressure recovery and leakage losses. This avoids the infeasibility or safety issues that arise in actual power plants due to neglecting physical boundaries, fully considering the physical characteristics of compressed gas energy storage and improving the engineering applicability and safety of the strategy.

[0022] Third, this invention uses the reduction of gas storage pressure to a safe threshold and the reduction of equivalent energy to a safe margin as trigger conditions for rolling correction, forming a multi-trigger mechanism together with traditional prediction bias criteria and risk change criteria. Since gas storage pressure is a slow variable unique to compressed gas storage, this mechanism can promptly trigger re-optimization when extreme weather causes a rapid drop in gas storage pressure, preventing the storage from being depleted. This solves the problem that fixed plans are difficult to adapt to extreme weather changes, establishing a rolling adaptive correction mechanism based on multiple criteria specific to compressed gas storage, thus enhancing the strategy's responsiveness to extreme weather changes.

[0023] Fourth, this invention establishes a dynamic trade-off mechanism between net load smoothing, renewable energy consumption, and emergency reserve by dynamically dividing capacity into four zones and rolling optimization of the emergency capacity shortage penalty term in the objective function. When the risk is low, more capacity is released to peak shaving tasks to improve the net load peak-valley difference and renewable energy consumption; when the risk increases, the emergency reserve ratio is automatically increased to ensure the reliability of system power supply, balancing peak shaving effect and emergency protection, and achieving a dynamic optimal trade-off among multiple objectives.

[0024] Fifth, the system of this invention includes a multi-power source collaborative verification module, which can combine thermal power, hydropower, demand response and electrochemical energy storage resources to carry out regional comprehensive regulation and verification, so that the peak-shaving path and reserve plan of compressed gas energy storage can meet the system supply and demand balance and reserve requirements on a larger scale, improve the operational resilience of high-proportion new energy power grid under extreme weather conditions, support regional multi-power source collaborative operation, and enhance the operational resilience of new power systems. Attached Figure Description

[0025] Figure 1 This is an overall flowchart of the method of the present invention.

[0026] Figure 2 This is a diagram showing the overall system structure of the present invention.

[0027] Figure 3 This is a schematic diagram of the dynamic capacity partitioning of the present invention.

[0028] Figure 4 This is a mapping diagram of thermal state and power grid capacity in this invention.

[0029] Figure 5 This is a schematic diagram illustrating the peak shaving and valley filling and emergency support principles of the present invention.

[0030] Figure 6 This is a schematic diagram illustrating the risk-adaptive emergency capacity reservation effect in this invention.

[0031] Figure 7 This is a comparison chart of net load before and after the optimization of peak-shaving paths under extreme weather windows according to the present invention. Detailed Implementation

[0032] To clearly illustrate the technical features of this solution, the following detailed implementation method will be used to explain the solution.

[0033] Example 1: This embodiment takes a large-scale advanced insulated compressed air energy storage power station with a rated power of 300MW and a rated capacity of 1800MWh as an example, which is connected to a provincial power grid and has a renewable energy installed capacity of about 65%. The dispatch cycle is 1 hour and the rolling window is 24 hours.

[0034] like Figure 1As shown, complete the following process step by step: Data Acquisition: Collect meteorological forecast data for the next 24 hours (temperature, wind speed, irradiance, humidity, precipitation warnings), historical load and renewable energy output data for the past 7 days, holiday information, and current status data for compressed air energy storage, including gas storage pressure, thermal storage temperature, and available power. Construct feature vector z. t Including: temperature T t Wind speed W t Irradiance G t Humidity H t Precipitation or disaster warning intensity R t Previous load 、 wind power output a moment ago The photovoltaic power output at the previous moment Time period markers (Including date, time period, and whether it is a holiday).

[0035] Joint prediction: Long Short-Term Memory (LSTM) network is used as the joint prediction model F θ Input historical sequence { } ， Output hourly load forecasts for the next 24 hours. Wind power forecast Photovoltaic forecast values and forecast uncertainty Set the confidence coefficient. =1.96 (corresponding to the 95% confidence interval), calculate the conservative upper limit of net load. .

[0036] Extreme Climate Risk Assessment: Setting Normal Ranges for Various Meteorological Elements: Normal Temperature Range (T) t nor Take the historical average for the same period, and the temperature limit deviation T lim =15℃; Normal wind speed range W t nor =4~8m / s, limit deviation W lim =10m / s; Normal irradiance range G t nor =400~600W / m 2 Limit deviation G lim =400W / m 2 Prediction uncertainty normalization factor σ lim =500MW; Supply-demand gap normalization factor D lim=1000MW. The weighting coefficients a1~a5 are determined using the entropy weighting method combined with historical extreme event data. In this example, a1=0.25, a2=0.15, a3=0.20, a4=0.25, a5=0.15. The limiting function is sat(x)=min(max(x,0),1). When a cold wave warning (temperature drop of more than 12℃) is issued within the next 6 hours, accompanied by low wind speeds (<3m / s), ρ t It can reach 0.7~0.9.

[0037] The weights a1 to a5 can be determined using the entropy weight method combined with historical extreme event data: collect records of extreme weather events over the past 5 years, extract features such as temperature deviation, wind speed deviation, irradiance deviation, standard deviation of prediction uncertainty, and supply-demand gap for each event, and construct an evaluation matrix; calculate the entropy value and weight of each indicator so that the weight reflects the actual contribution of each factor to the system risk. Alternatively, expert scoring or machine learning feature importance ranking can be used for determination, both of which do not deviate from the core idea of ​​this invention.

[0038] Compressed gas energy storage capacity assessment: Reference Figure 4 The diagram showing the thermal state of compressed gas storage and its mapping to grid capacity is used to obtain the current gas storage pressure p. t =7.2MPa (rated range 4~8MPa), thermal storage temperature 280℃, usable power 300MW, ramp rate 60MW / min, continuous support time 6h (full power). Calculate the usable regulation capacity C. t ava =1800MWh (equivalent energy capacity). Set the safe gas storage pressure threshold p. safe =5.0MPa, safety energy margin E safe =200MWh.

[0039] Dynamic capacity partitioning: such as Figure 3 As shown, the baseline allocation coefficients are set to α0 = 0.3, α1 = 0.4, and α2 = 0.3. The baseline net load N... t base =1500MW. If ρ during a certain period t =0.8, =2200MW, then =1800×[0.3+0.4×0.8+0.3×(2200-1500) / 300]=1800×[0.3+0.32+0.7]=1800×1.32=2376MWh, exceeding the total capacity. = =1800MWh, =0, meaning all capacity is reserved for emergency use. In practical applications, this is ensured through calibration from α0 to α2. No more than When the calculated value exceeds the limit, the upper limit is taken.

[0040] Peak shaving path optimization: Construct an objective function J with weights w1=10, w2=50, w3=100, w4=1, and w5=5. Employ a model predictive control framework and call a commercial solver (such as Gurobi) to solve a mixed-integer quadratic programming problem, obtaining hourly values ​​for the next 24 hours. , and backup / reserve plans. Backup requirements. , where R0=100MW, β1=0.3, β2=0.2, β3=0.5.

[0041] Multi-power source coordination verification: The peak-shaving path of compressed gas energy storage is integrated with the power plans of thermal power, hydropower, demand response and other energy storage resources for regional comprehensive regulation verification to ensure that the regional comprehensive regulation capacity meets the requirements for peak shaving and reserve.

[0042] Rolling correction: The actual net load is updated every hour. Calculate using the latest forecast data ,set up =200MW, =0.2. If ≥200MW or ≥0.2 or ≤200MWh or If the pressure is ≤5.0MPa, a re-optimization will be triggered, updating only the strategy within the future rolling window, while the execution period remains fixed.

[0043] Output the current time period's compressed gas storage charging and discharging power command, emergency capacity reserve ratio, risk warning level (green / yellow / orange / red), and strategy explanation text.

[0044] To visually demonstrate the technical effects of this invention, a conventional peak-shaving strategy is set as a baseline for comparison. The conventional strategy employs a fixed day-ahead plan and risk-free emergency reserve, i.e., it does not consider p. t Dynamic adjustments to capacity partitions are not configured. The rolling correction trigger condition reserves only a small amount of spare capacity at a fixed ratio.

[0045] Under this baseline strategy, compressed gas storage may experience excessive discharge due to routine peak-shaving tasks before the arrival of extreme weather windows, leading to high-risk periods of gas storage pressure p. t and equivalent energy E t Rapidly dropping below the safety threshold results in a loss of emergency support capabilities. This invention, however, utilizes ρ... t drive Dynamic enhancement allows for proactive compression before extreme weather windows. ,reserve and ,make and It is always maintained above the safety threshold, while taking into account both net load smoothing and renewable energy consumption.

[0046] Construct a 72-hour rolling scenario: the first 24 hours are under normal weather conditions (ρ t ≈0.1~0.2), with an extreme weather window of 30 to 48 hours (cold wave + low radiation, ρ t The concentration rose to 0.7-0.9, then returned to normal within 24 hours. The above-mentioned conventional peak-shaving strategy was used as a control.

[0047] Define the net load peak-to-valley difference improvement rate and emergency capacity satisfaction rate As an evaluation indicator: Let the net load before optimization be The optimized net load is ,but ; Let the emergency capacity requirement for time period t be... The actual reserved capacity is ,but .

[0048] Reference Figure 5 This embodiment uses visualization to demonstrate the dual benefits of compressed gas energy storage systems participating in grid regulation: 1. Peak shaving and valley filling to smooth net load fluctuations Without compressed air energy storage, the net load of the power grid exhibits drastic fluctuations (solid black line). By implementing the optimization strategy described in this invention, during the early morning off-peak period (left side of the figure), the system controls the compressed air energy storage to absorb and compress excess power from the grid, slightly boosting the net load curve (corresponding to the area where the off-peak is filled in the figure), effectively avoiding frequent deep peak shaving by conventional units; during the afternoon peak load period (middle area of ​​the figure), the system controls the compressed air energy storage to release stored high-pressure air for power generation, reducing the steep peak (corresponding to the area where the peak is reduced in the figure), significantly reducing the peak net load, thereby greatly improving the stability of the power grid load.

[0049] 2. Dynamically reserve emergency capacity during extreme periods The optimization path of this invention does not solely pursue the ultimate smoothing effect, but also takes into account the system's safety margin. As shown in the right-hand area of ​​the figure, when facing extreme weather or periods of high uncertainty, the optimized net load curve (dashed line) does not drop indefinitely, but always remains below the emergency capacity reserve curve (baseline). The emergency capacity reserve marked in the figure intuitively demonstrates that the system still has sufficient backup support capacity when facing sudden load changes, verifying the advantages of this invention in balancing economy and safety.

[0050] Reference Figure 6 The diagram showing the linkage between risk index, emergency capacity reserve, and compressor storage energy status within extreme weather windows illustrates the changes in risk index ρ during normal weather periods (0-30h and 50-70h). t Maintaining a low level, the system primarily allocates available compressed air storage capacity to routine peak shaving, with emergency capacity at a baseline level. However, when the simulation enters the set extreme weather window (30-50 hours, blue shaded area), the risk index surges, and the system immediately triggers an adaptive response mechanism. The emergency capacity reserve ratio rapidly increases, ensuring sufficient backup support capacity during periods of high uncertainty, thus verifying the dynamic response characteristics of the capacity dynamic zoning module. Simultaneously, the equivalent energy state E of compressed air storage... t Although the discharge rate decreases slightly within the extreme window due to peak-shaving discharge, it still remains within the safe threshold E. safe The above demonstrates the effectiveness of this invention in balancing peak shaving and emergency response.

[0051] Reference Figure 7 The chart showing the net load comparison before and after peak-shaving path optimization under extreme weather windows illustrates the hourly charge-discharge plan for compressed gas storage over the next 72 hours after solving the mixed-integer linear programming model. This example demonstrates the typical peak-shaving effect within an extreme weather window (30-50 hours, light blue shaded area): Flexible regulation during normal periods (0~30h and 50~70h): During normal weather periods, the system mainly utilizes compressed air energy storage for basic peak shaving and valley filling. As shown by the orange curve in the figure, during low load periods (such as 10~20h), compressed air energy storage absorbs excess power from the grid and compresses it, slightly boosting the net load and avoiding frequent deep peak shaving by conventional units; during high load periods (such as 60~70h), compressed air energy storage releases power, effectively reducing the peak net load.

[0052] Extreme weather window emergency protection (30~50h): When the simulation enters the extreme weather window, the original net load curve (blue solid line) shows a steep peak (close to 550MW) around 40h. At this time, based on the emergency capacity boundary reserved by the dynamic capacity zoning, the compressed gas storage does not blindly pursue the ultimate smoothing effect. Instead, under the premise of meeting the high reserve requirements of the system, it significantly reduces the peak-to-valley difference of the system and retains sufficient buffer margin for load changes under extreme weather conditions through precise power regulation (suppressing the net load at around 450MW).

[0053] The simulation results are as follows:

[0054] The results show that the present invention maintains a high emergency capacity satisfaction rate (0.97, compared to only 0.62 for conventional strategies) and gas storage pressure margin (minimum 5.4 MPa, higher than the safety threshold of 5.0 MPa) during extreme weather windows. At the same time, the net load peak-valley difference improvement rate is increased by about 4 percentage points, and the amount of renewable energy curtailment is reduced by about 60%, which verifies the effectiveness of the present invention in balancing peak shaving effect and emergency protection under extreme climate conditions.

[0055] In the above embodiments, the joint prediction model F θ Taking LSTM as an example, those skilled in the art will understand that F θ It can also be replaced by random forest, gradient boosting tree, Transformer, or mechanistic data fusion model; extreme climate risk index ρ t Taking weighted summation plus limiting as an example, fuzzy evaluation, Bayesian networks, or scenario probability methods can also be used to construct the rolling peak shaving path optimization, taking model predictive control plus MILP solution as an example, stochastic optimization, robust optimization, or heuristic intelligent optimization methods can also be used to achieve it; in addition to thermal power, hydropower, demand response, and electrochemical energy storage, multi-power source collaboration objects can also be extended to virtual power plants, electric vehicle aggregated resources, etc.

[0056] Example 2: Reference Figure 2 A system employing an adaptive planning method for compressed gas storage peak-shaving paths and an emergency capacity reservation method for extreme climate scenarios includes: The data acquisition module is used to collect data on weather forecasts, load, new energy output, compressed gas storage status, and grid constraints. The joint forecasting module is used to perform joint forecasting and output load forecast values. Wind power forecast Photovoltaic forecast values and forecast uncertainty ; Extreme climate risk identification module, used to perform ρ t calculate; The compressed gas energy storage status assessment module is used to obtain gas storage pressure, thermal storage status, available power, ramp-up capability, and continuous support time. The capacity dynamic partitioning module is used to perform... Four-zone division and calculate; The peak shaving path optimization module is used to perform rolling robust optimization and output charging power. Discharge power and backup plans; The multi-power source collaborative verification module is used to combine the regulation power of thermal power, hydropower, demand response and electrochemical energy storage, and to perform regional comprehensive regulation and verification of the peak-shaving path and reserve plan of compressed gas energy storage. The rolling correction module is used to perform triggered re-optimization. The operation strategy output module is used to output peak shaving paths, emergency capacity reservation plans, early warning levels, and strategy explanations.

[0057] The capacity dynamic zoning module uses the ρ output by the extreme climate risk identification module. t and the output of the joint prediction module Calculate emergency capacity for extreme weather The calculation results are then output to the peak shaving path optimization module; the triggering conditions for the rolling correction module include... ,in and The minimum available pressure of the compressed gas storage chamber and the energy required for emergency support are pre-calibrated and stored in the compressed gas storage status assessment module.

[0058] The system in this embodiment is deployed in the energy management system of the provincial dispatch center. The data acquisition module collects data every 15 minutes via the SCADA interface. The joint prediction module is deployed on a GPU server and executes LSTM inference every hour. The extreme climate risk identification module receives early warning information issued by the meteorological bureau and automatically updates p. t The capacity dynamic partitioning module is based on ρ t and The system calculates the capacity of each zone in real time. The peak-shaving path optimization module calls the optimization solver to complete 24-hour rolling optimization within 30 seconds. The multi-power source collaborative verification module interacts with the thermal power automatic generation control system, hydropower dispatching system, and demand response management platform through the dispatch data network to obtain the adjustable output range and current instructions of each conventional power source, and performs regional comprehensive adjustment and verification of the peak-shaving path and reserve plan of the compressed gas storage. The rolling correction module monitors deviations and automatically starts re-optimization when the trigger conditions are met. The operation strategy output module sends instructions to the station control system of the compressed gas storage power station.

[0059] This invention can be deployed in dispatch automation systems, energy storage power station energy management systems, or virtual power plant control platforms. For the dispatch side, the compressed gas storage charging and discharging paths, emergency capacity reservation curves, and risk warning levels output by this invention can be used as auxiliary decision-making results for intraday rolling plans. For the power station side, it can be used as an upper-level strategy in the energy storage power station energy management system, issuing power references and capacity reservation constraints to the compressor, expander, and grid-connected control system.

[0060] In engineering implementation, the joint forecasting module can be updated every 15 minutes to 1 hour, the rolling optimization module can use a forecast window of 24 hours or 48 hours in the future, and the execution layer only issues instructions for the current control cycle or a few short cycles, thus taking into account both the foresight of the plan and the stability of the execution.

[0061] When the dispatching agency issues an extreme weather warning or a system reserve shortage warning, this invention can automatically switch to a conservative operation mode to improve efficiency. and Once the warning is lifted and the actual operational deviations have decreased, the system will gradually release emergency capacity and resume regular peak shaving or priority strategies for renewable energy consumption.

[0062] For regional integrated regulation scenarios involving multiple compressed air energy storage power plants or compressed air energy storage + thermal power + hydropower + demand response, it can be further expanded into a distributed collaborative optimization form, where the regional control center allocates peak-shaving and reserve tasks for each resource, and each power plant performs local optimization based on its own thermal state and safety margin.

[0063] Example 3: A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an adaptive planning method for compressed gas storage peak shaving paths and an emergency capacity reservation method for extreme climate scenarios.

[0064] Of course, the above description is not limited to the examples above. Technical features not described in this invention can be implemented by or using existing technology, and will not be repeated here. The above embodiments and drawings are only used to illustrate the technical solutions of this invention and are not intended to limit this invention. This invention has been described in detail with reference to preferred embodiments. Those skilled in the art should understand that any changes, modifications, additions or substitutions made by those skilled in the art within the scope of this invention do not depart from the spirit of this invention and should also fall within the scope of protection of the claims of this invention.

Claims

1. A method for adaptive planning of peak-shaving paths and emergency capacity reservation for compressed gas storage in extreme climate scenarios, characterized in that: Includes the following steps: Collect meteorological forecast data, historical load data, renewable energy output data, and compressed gas storage operation status data for the target time period, and construct the feature vector z for time period t. t eigenvector z t It should include at least meteorological elements, historical loads, historical renewable energy output, and calendar characteristics; Through the joint prediction model F θ Output load forecast values ​​for each time period within the future rolling window. Wind power forecast Photovoltaic forecast values and forecast uncertainty And calculate the upper limit of the conservative net load. , Confidence coefficient; Constructing an extreme climate risk index : ； Among them, T t W t G t These are temperature, wind speed, and irradiance, respectively. t To predict the supply-demand gap, a1~a5 are weighting coefficients, and sat(·) is the limiting function; Gas storage pressure p based on compressed gas energy storage t Thermal storage status, climbing ability, and continuous support time; calculation of available adjustment capacity for time period t. and will Dynamically divided into peak-shaving capacity Frequency modulation reserve capacity Extreme weather emergency capacity and safety minimum capacity ,in: ； α0~α2 are the distribution coefficients. Based on the baseline net load, This is the rated power of compressed air energy storage; A rolling robust optimization model is constructed with the objectives of smoothing net load, reducing renewable energy curtailment, ensuring sufficient emergency capacity, and minimizing operating costs to generate the charging power of compressed gas storage. Discharge power and a backup plan, and meet the backup constraint; ； in, As a backup for conventional units, Reserved for demand response; Backup required for the system under extreme weather conditions: ； β1~β3 are reserve coefficients; Let the actual net load be N. t real Define prediction bias The equivalent energy state of compressed gas energy storage is denoted as E. t The gas storage pressure is denoted as p. t ;ε N ε ρ These are the net load deviation threshold and the risk change threshold, respectively. safe p safe These are the preset safe energy margin and safe gas storage pressure threshold, respectively; Forecast data and extreme climate risk index are updated on a rolling basis. And compressed gas energy storage state, when satisfied Time-triggered re-optimization; Output peak shaving routes, emergency capacity reservation plans, and risk warning levels.

2. The adaptive planning method for compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios as described in claim 1, characterized in that: Feature vector z t Including temperature T t Wind speed W t Irradiance G t Humidity H t Precipitation or disaster warning intensity R t Historical load Historical wind power output Historical photovoltaic power output and features of date, time period and holidays. .

3. The adaptive planning method for compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios as described in claim 1, characterized in that: The equation of state for compressed gas energy storage includes the equivalent energy state E. t and gas storage pressure p t : ； ； in and These represent charging and discharging efficiencies, respectively. , , This is the pressure dynamic coefficient. Due to environmental pressures, This refers to gas leakage from the gas storage chamber. The operational constraints for compressed gas energy storage are: ； ； ； in , These are the maximum charging power and the maximum discharging power, respectively. , These are the upper and lower limits of the equivalent energy, respectively. , These are the upper and lower limits of the gas storage pressure. , These represent the charging and discharging ramp rates, respectively.

4. The adaptive planning method for compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios as described in claim 1, characterized in that: The objective function for rolling robust optimization is: ； in For real-time net load, Based on net load, For the abandoned electricity of new energy sources, As a penalty for insufficient emergency capacity, Δ The change in compressed gas energy storage power is represented by w1~w5, which are the weights.

5. The adaptive planning method for compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios as described in claim 1, characterized in that: When the risk of extreme weather increases Increase or conservatively limit the upper limit of net load Exceeding the baseline net load At the same time, extreme weather emergency capacity Automatically increase and peak shaving capacity The corresponding compression will reduce the gas storage pressure p before the arrival of the high-risk period. t and safety minimum capacity Maintain a margin; release gradually once the risk is eliminated. Resume regular peak shaving.

6. The adaptive planning method for compressed gas storage peak-shaving paths and emergency capacity reservation for extreme climate scenarios as described in claim 1, characterized in that: Joint prediction model F θ The model can be one of the following: random forest, gradient boosting tree, long short-term memory network, Transformer, or mechanistic data fusion model. The rolling robust optimization adopts a model prediction control framework with a rolling window of 24h or 48h and a control period of 1h or 15min.

7. A system for adaptive planning of compressed gas storage peak-shaving paths and emergency capacity reservation method for extreme climate scenarios as described in any one of claims 1 to 6, characterized in that: include: The data acquisition module is used to collect data on weather forecasts, load, new energy output, compressed gas storage status, and grid constraints. The joint forecasting module is used to perform joint forecasting and output load forecast values. Wind power forecast Photovoltaic forecast values and forecast uncertainty ; Extreme climate risk identification module, used to perform ρ t calculate; The compressed gas energy storage status assessment module is used to obtain gas storage pressure, thermal storage status, available power, ramp-up capability, and continuous support time. The capacity dynamic partitioning module is used to perform... Four-zone division and calculate; The peak shaving path optimization module is used to perform rolling robust optimization and output charging power. Discharge power and backup plans; The multi-power source collaborative verification module is used to combine the regulation power of thermal power, hydropower, demand response and electrochemical energy storage, and to perform regional comprehensive regulation and verification of the peak-shaving path and reserve plan of compressed gas energy storage. The rolling correction module is used to perform triggered re-optimization. The operation strategy output module is used to output peak shaving paths, emergency capacity reservation plans, early warning levels, and strategy explanations.

8. The system according to claim 7, characterized in that: The capacity dynamic zoning module uses the ρ output by the extreme climate risk identification module. t and the output of the joint prediction module Calculate emergency capacity for extreme weather The calculation results are then output to the peak shaving path optimization module. The triggering conditions for the scroll correction module include ,in and The minimum available pressure of the compressed gas storage chamber and the energy required for emergency support are pre-calibrated and stored in the compressed gas storage status assessment module.

9. A computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method of any one of claims 1 to 6.

Images