Multi-time-scale toughness scheduling strategy for multi-energy complementary system under extreme high temperature condition

Abstract

The invention relates to a multi-time-scale toughness scheduling strategy of a multi-energy complementary system under an extreme high temperature condition in optimal scheduling of a power system. In order to solve the problems of load rising caused by high temperature, derating of a source network and difficulty in guaranteeing power supply safety by traditional economic dispatching, a source-network-load-storage temperature effect model of photovoltaic, thermal power, a power transmission transformer and a load is constructed, and a day-ahead, day-intraday and real-time three-layer collaborative optimization framework is embedded; weighted unsupplied electric quantity is used as a toughness index, node vulnerability and load grade weight are combined, dictionary order optimization is adopted before the day, prediction deviation is corrected in a rolling mode within the day, toughness-oriented model prediction control and rolling supply stop window constraint are adopted in real time, and conventional unit, energy storage and layered demand response are cooperatively scheduled. Therefore, load loss is reduced and new energy is abandoned in an extreme high-temperature scene, and the power supply guarantee capability of key nodes and important loads and the overall toughness and economical efficiency of the system are improved.

Description

Technical Field

[0001] This invention belongs to the field of power system operation and control technology, specifically relating to a multi-timescale resilient dispatch strategy for multi-energy complementary power systems facing extreme high-temperature scenarios. It is applicable to integrated source-grid-load-storage operation scenarios that simultaneously include multiple power sources such as wind power, photovoltaic, hydropower, thermal power, and energy storage connected to the power grid. Background Technology

[0002] In recent years, global warming has continued, leading to a significant increase in the frequency of extreme heat waves and heatwave events. Widespread and sustained high temperatures cause a sharp increase in temperature control loads on residential and commercial buildings, resulting in a surge in peak system loads and a widening peak-to-valley difference. Simultaneously, they reduce the efficiency of cold-end power units, lower the thermal capacity of transmission lines and transformers, and cause photovoltaic module efficiency to decline with increasing temperature. This significantly compresses the overall power system's supply capacity and safety margin. Under extreme heat conditions, conventional economic-driven dispatching methods struggle to reflect the temperature degradation characteristics of available equipment capacity in a timely manner, easily underestimating system operational risks and potentially triggering localized load shedding and widespread power curtailment.

[0003] To enhance the absorption capacity and operational flexibility of renewable energy, multi-energy complementary systems, such as wind-solar-hydro-thermal-storage systems, are widely used in power systems. Through source-side complementarity and inter-period energy transfer via energy storage, multi-energy complementary systems can achieve peak shaving and valley filling, smooth fluctuations, and improve power supply reliability under normal conditions. However, existing dispatching methods often assume conventional meteorological conditions. The impact of extreme high temperatures on the efficiency and capacity of thermal power units, the upper limit of photovoltaic output, the upper limit of transmission line thermal stability, and the allowable load factor of transformers is often only handled with simple reductions or empirical coefficients, lacking a unified modeling of time-varying constraints from ambient temperature to equipment. This leads to a deviation between the feasible region of the dispatching solution and the actual operating boundary under extreme high-temperature scenarios. On the other hand, the power grid supply and demand contradiction is prominent under extreme high-temperature scenarios, inevitably requiring demand-side response and necessary load reduction. Existing research mainly aims to reduce total load or lower overall costs, lacking differentiated consideration of the vulnerability of different nodes and the importance of different load levels. It is difficult to provide precise protection for critical loads and critical channels, and there is a lack of constraints on the fairness of power outage timing. This can easily lead to problems such as prolonged and continuous power outages at some nodes, which is not conducive to improving the overall resilience of the system.

[0004] With the increasing penetration of stochastic power sources such as wind and solar power, the operating characteristics of power systems exhibit significant multi-timescale features. Day-ahead, intraday, and real-time dispatch are nested and mutually influential across time scales. Existing dispatch strategies typically focus on economic optimization, making it difficult to establish resilience benchmarks and transmission mechanisms across time scales under extreme high-temperature scenarios. On the one hand, the day-ahead stage lacks explicit resilience targets, failing to provide quantifiable resilience constraints for subsequent stages. On the other hand, while rolling optimization and model predictive control can be introduced to offset some prediction errors in the intraday and real-time stages, if not coupled with resilience targets and relying solely on local economic optimization, the system's overall resilience to high-temperature disturbances may still be weakened after multiple rolling corrections.

[0005] Therefore, it is necessary to construct a resilient scheduling strategy that is oriented towards extreme high-temperature conditions, can uniformly characterize the temperature effects of the source-grid-load-storage system, and coordinates multiple time scales. This strategy should take into account factors such as environmental temperature-driven equipment performance degradation, node vulnerability, and load levels, and improve the operational resilience and economy of multi-energy complementary systems in extreme high-temperature scenarios while ensuring the power supply safety of critical loads and critical channels. Summary of the Invention

[0006] The purpose of this invention is to address the problems of equipment capacity derating, load surges, and significant random fluctuations in source and load in multi-energy complementary power systems under extreme high-temperature conditions. It proposes a multi-timescale resilient scheduling strategy, based on a unified temperature effect model, constructing a three-layer collaborative resilient-oriented scheduling framework that integrates day-ahead, intraday, and real-time operations. This framework introduces a weighted unsupplied power quantity index based on node vulnerability and load level, as well as rolling outage window constraints, to achieve differentiated protection and fair control of outage timing for critical loads and critical channels. This enables synergistic optimization of system economy and resilience under extreme high-temperature scenarios.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] 1. Temperature Effect Modeling. Based on historical ambient temperature data and meteorological forecasts for the target area, temperature corrections are applied to the operating characteristics of various power sources, networks, and loads in the multi-energy complementary system to establish a unified source-grid-load-storage temperature effect model. Specifically, this includes: adjusting the upper limit of photovoltaic unit output at different times based on the relationship between ambient temperature and photovoltaic cell temperature to obtain the available photovoltaic capacity varying with temperature; establishing a piecewise linear relationship between unit efficiency and available capacity and ambient temperature based on the efficiency degradation and capacity derating characteristics of thermal power units under cold-end limited conditions to obtain the multiplier of available capacity for thermal power units at different times; establishing a temperature-sensitive load model based on the impact of ambient temperature on temperature-controlled loads, introducing a temperature gain factor to characterize load increases caused by high temperatures; determining the allowable load factor characterizing transformer capacity derating based on a transformer hotspot temperature calculation model, mapping ambient temperature to the time-varying capacity upper limit of the transformer; and calculating the dynamic thermal stability current upper limit of transmission lines under different ambient temperatures based on the thermal balance relationship of transmission line conductors, mapping it to the branch power transmission upper limit to form a network transmission capacity constraint considering the impact of extreme high temperatures.

[0009] 2. Construction of a multi-timescale scheduling framework. The scheduling time is divided into three stages: day-ahead, intraday, and real-time. Each stage shares a unified source-grid-load-storage temperature effect model and equipment operation constraints. By using memory variables to transfer unit start-up and shutdown status, energy storage charge status, rolling outage window status, and weighted unsupplied power resilience benchmark between different stages, state inheritance and resilience constraint transfer are achieved across multiple timescales.

[0010] 3. Day-ahead Dispatch. Using an hourly time resolution, and under the influence of node vulnerability weights and load level weights, the first-stage objective is the system-weighted unsupplied power volume. A lexicographical optimization approach is employed to first minimize the weighted unsupplied power volume over the entire dispatch cycle, yielding a resilience benchmark solution considering the impact of extreme high temperatures. Under the constraint that the system-weighted unsupplied power volume is not inferior to this resilience benchmark solution, the second-stage objective is the comprehensive benefit comprised of generation revenue, energy storage operating costs, demand response subsidy costs, renewable energy curtailment penalty costs, and direct load shedding penalty costs. This involves optimizing the start-up and shutdown status of thermal power units, the pumping and generation sequence of pumped storage units, the electrochemical energy storage charging and discharging plan, and different types of demand response call schemes. Discrete decision variables and the resilience benchmark are then passed as memory variables to the intraday stage.

[0011] 4. Intraday Rolling Scheduling. Using a time resolution of less than 15 minutes and a rolling outlook window of several hours, at the beginning of each rolling window, the system reads the memory variables passed from the previous day's stage, as well as the updated energy storage state of charge and rolling shutdown window status from the previous window. Based on the updated renewable energy output and load forecasts, rolling optimization is performed on the active power output of thermal power units, the charging and discharging power of electrochemical energy storage, the output of pumped storage units, and the amount of incentive-based demand response calls. While maintaining the unit start-up and shutdown status unchanged and ensuring that some demand response schemes are consistent with the previous day's stage, the system constrains the weighted unsupplied power at each node within the current window to not exceed the resilience benchmark for the corresponding time period, and guarantees at least one hour of full power supply within any consecutive rolling shutdown window of a preset number of hours. The optimized energy storage state of charge, rolling shutdown window status, and weighted unsupplied power at each node are then passed as updated memory variables to the next intraday rolling window or real-time stage.

[0012] 5. Real-time Resilience-Oriented Scheduling. Using time steps of less than 5 minutes and a prediction domain length of several time steps, the system state variables and ambient temperature are updated based on real-time measurement data within each real-time rolling window. Ultra-short-term forecasts are used to obtain the net load and renewable energy output trajectories for the next few steps. With the real-time operating cost and weighted sum of unsupplied power as the objective, and under the conditions of satisfying the physical constraints of source-grid-load-storage, temperature effect constraints, energy storage state of charge constraints, and rolling outage window constraints, the control sequence for the marginal output of thermal power units, electrochemical energy storage output, and incentive-based demand response calls within the prediction domain is optimized. Only the first step of this control sequence is implemented, updating the system state and memory variables, and advancing the time by one scheduling step. This optimization process is repeated to achieve real-time rolling feedback correction for multi-timescale resilient scheduling.

[0013] 6. Dispatch Instruction Generation. Based on the multi-timescale dispatch results obtained at the day-ahead, intraday, and real-time stages, generator output plans, energy storage charging and discharging instructions, demand response call strategies, and necessary load reduction and power restoration sequences for extreme high-temperature scenarios are generated and issued to the corresponding generation, energy storage, and load-side execution devices to achieve multi-timescale resilient dispatch of the multi-energy complementary system under extreme high-temperature conditions. Based on the above method, preferably, the construction of the system's weighted unpowered quantity includes three aspects: node vulnerability, load level, and unpowered quantity. First, a set of nodes and a set of load levels for the multi-energy complementary system are established. The vulnerability of each node is assessed using the node power flow betweenness index to obtain the vulnerability weight of each node. Then, the total load of each node is divided into at least four levels according to importance, and a corresponding load level weight is set for each level. Finally, the unpowered quantity of each node at each level is weighted and summed according to the vulnerability weight and load level weight to obtain the objective function of the system's weighted unpowered quantity. One implementation of this function can be expressed as:

[0014]

[0015] In the formula, The weighted amount of power not supplied to the system within a scheduling cycle. For the set of scheduling periods, For the node set index, For load level set index, w i Let π be the vulnerability weight of node i. j For load level weighting, To reduce the load, Δt DA Let be the scheduling time interval. By minimizing the weighted unpowered quantity objective function shown in equation (1), power supply to highly vulnerable nodes and high-level loads can be prioritized during the optimization process.

[0016] Preferably, the rolling power outage window constraint is implemented by introducing a full power supply indicator variable for each load node. When the full power supply indicator variable is 1, the amount of power outage for the corresponding time period of that node is constrained to be zero; for any consecutive preset number of hours of sliding time window, the sum of the full power supply indicator variables within that window is constrained to be no less than 1, thereby avoiding long-term continuous power outages for load nodes and achieving fair control of power outage timing. Specifically, for any load node i and any time period τ as the starting point, the following constraints can be set:

[0017]

[0018] In the formula, Let be the indicator variable for the full power supply of node i during time period t, when the node achieves full power supply during that time period. otherwise This formula guarantees that there is at least one period of complete power supply within any consecutive time period, thereby avoiding prolonged continuous power outages.

[0019] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the above-described method for implementing a multi-timescale resilient scheduling strategy for a multi-energy complementary system under extreme high-temperature conditions are performed.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] (1) Based on ambient temperature and meteorological forecast data, this invention constructs a unified temperature effect model for photovoltaic units, thermal power units, transmission lines, transformers and temperature-sensitive loads, and explicitly maps the performance degradation of the entire process of source-grid-load-storage by extreme high temperature to the time-varying output limit and capacity coefficient in the scheduling constraints, thereby improving the accuracy of the scheduling model in depicting the actual operating boundary.

[0022] (2) This invention introduces vulnerability weights and load level weights based on node power flow betweenness to construct a system weighted unpowered quantity index. The weighted unpowered quantity is minimized in the day-ahead stage using lexicographical optimization to establish a quantitative resilience benchmark. Economic adjustments are made under the constraints of this resilience benchmark in the intraday and real-time stages, thereby avoiding the problem of the resilience level being repeatedly eroded during multiple rolling optimization processes.

[0023] (3) The present invention implements time constraints on unplanned power outages of each load node by using rolling power outage window constraints, ensuring that at least one period of time within any consecutive preset number of hours is fully powered, effectively suppressing long-term continuous power outages, and taking into account the fairness of power outages among different regions and users.

[0024] (4) This invention constructs a multi-timescale scheduling framework consisting of day-ahead coordination, intraday rolling and real-time feedback correction. By utilizing the response characteristics of energy storage systems and hierarchical demand response at different timescales, it realizes the coordinated utilization of flexible resources on the source side, grid side and load side, and improves the supply and demand matching capability and operational resilience of multi-energy complementary systems under extreme high temperature conditions.

[0025] (5) The present invention adopts a resilience-oriented model predictive control method to achieve rolling optimization and feedback correction in the real-time stage. Under the premise of ensuring the physical constraints of source-grid-load-storage and temperature effect constraints, it can absorb the ultra-short-term random disturbances of wind power, photovoltaic and load, and reduce the risk of load loss and the level of new energy curtailment under extreme high temperature scenarios. Attached Figure Description

[0026] To more clearly illustrate the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0027] Figure 1 This is a schematic diagram of node vulnerability assessment results provided in an embodiment of the present invention;

[0028] Figure 2 This is a schematic diagram of the overall framework for multi-timescale resilient scheduling provided in an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of the multi-timescale resilient scheduling execution process provided in an embodiment of the present invention;

[0030] Figure 4 This is a schematic diagram of the typical computational system topology and equipment configuration provided in an embodiment of the present invention;

[0031] Figure 5 This is a schematic diagram of the system power balance comparison curve provided in an embodiment of the present invention. Detailed Implementation

[0032] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the following embodiments are only for illustrating the present invention and are not intended to limit the present invention; various modifications or equivalent substitutions can be made to the following embodiments by those skilled in the art without departing from the spirit and substance of the present invention, and all such modifications or substitutions should fall within the protection scope of the present invention.

[0033] Example 1: Modeling of the temperature effect of the source-grid-load-storage system

[0034] This embodiment targets a multi-energy complementary system that includes wind power, photovoltaic power, hydropower, thermal power, pumped storage, electrochemical energy storage, and loads of varying importance levels. Based on historical ambient temperature data and meteorological forecast information, a unified source-grid-load-storage temperature effect model is established.

[0035] 1. Modeling of temperature effect in photovoltaic units

[0036] Under extreme high-temperature conditions, the temperature of photovoltaic module cells rises, leading to a decrease in their photoelectric conversion efficiency. Considering the linear relationship between ambient temperature and cell temperature, the upper limit of usable output of a photovoltaic unit can be expressed as:

[0037] P PV =P STC ·η std [1-γ PV ·(T cell -T ref,PV (3)

[0038] In the formula, P PV P represents the available output of the photovoltaic array under current conditions. STC η is the rated output under standard test conditions. std For the inverter's equivalent efficiency under standard conditions, γ PV T is the temperature power coefficient of a photovoltaic module. cell For the cell temperature, T ref,PV This is the reference temperature for photovoltaics.

[0039] 2. Temperature effect modeling of thermal power units

[0040] Under conditions of increased cooling water temperature and limited cold-end capacity, the condenser vacuum of thermal power units deteriorates, leading to a decrease in unit thermal efficiency and a reduction in available output. This embodiment uses a piecewise linear model to represent the changes in unit efficiency and available capacity with ambient temperature. For example, the equivalent power generation efficiency of the unit can be expressed as:

[0041] η TH,t =η N -α TH [T amb,t -T thr,TH ] + ,0<η TH,t≤1 (4)

[0042] In the formula, η TH,t The current ambient temperature T amb,t The equivalent efficiency of the unit, η N α is the unit efficiency under rated operating conditions. TH T is the sensitivity coefficient of efficiency to ambient temperature. thr,TH The threshold temperature at which power generation efficiency begins to be affected by temperature, [ ] + This is a positive multiplier operator. Thermal power units can use the capacity multiplier φ. t It can be represented as:

[0043] φ t =min{φ max ,φ max -η thr T amb,t},≥ t ≥0 (5)

[0044] In the formula, ≥ max η represents the upper limit of the available capacity of thermal power units. thr T is the temperature sensitivity coefficient of thermal power units. amb,t Given the current ambient temperature, the upper limit of active power output of thermal power units during dispatching is determined by their rated capacity and φ. t The product is determined.

[0045] 3. Modeling of temperature-sensitive loads

[0046] Temperature-controlled loads, such as air conditioning, increase significantly under high-temperature conditions. To reflect the impact of ambient temperature on net load, this embodiment uses a temperature-sensitive load model:

[0047] L t =L base +β load [T amb,t -T thr,load ] + (6)

[0048] In the formula, L t Let L be the load level for time period t. base For the baseline load, β load T is the sensitivity coefficient of the load to temperature. amb,t For ambient temperature, T thr,load This is the threshold temperature at which temperature-sensitive loads begin to increase.

[0049] 4. Transformer temperature effect modeling

[0050] Based on the transformer hotspot temperature calculation model, this embodiment uses the allowable load factor to characterize the transformer capacity derating. Let κ... TR,t The allowable load factor of the transformer during time period t can be expressed as:

[0051] κ TR,t =1-β TR (T amb,t -T ref,TR ),0≤κ TR,t ≤1 (7)

[0052] In the formula, β TR T represents the derating slope of the transformer capacity with respect to ambient temperature. ref,TR This refers to the rated operating ambient temperature. In the dispatch model, the upper limit of active power transmission of the transformer is determined by the rated capacity and κ. TR,t The product is given.

[0053] 5. Modeling of temperature effects in transmission lines

[0054] The current that a transmission line can carry is affected by the temperature rise of the conductors and the ambient temperature. This embodiment maps the ambient temperature to the upper limit of the dynamic thermal stability current of the line based on the conductor's thermal balance relationship, and then further converts it into the upper limit of the active power transmission of the branch. Under a linear approximation, it can be written as:

[0055]

[0056] In the formula, This represents the upper limit of dynamic thermal stability that the line can transmit. β represents the rated transmission capacity under reference conditions. line T is the derating factor for the transmission capacity of a line due to temperature. amb,t For ambient temperature, T ref,line The rated operating temperature of the transmission line;

[0057] The above modeling method enables unified temperature effect modeling of photovoltaic units, thermal power units, transmission lines, transformers and load demand under extreme high temperature conditions, providing a foundation for subsequent multi-timescale resilient scheduling.

[0058] Example 2: Node Vulnerability Assessment and Weighted Unpowered Quantity Construction

[0059] This embodiment assesses system node vulnerability based on the node power flow betweenness index and constructs a weighted unpowered quantity index that takes into account both node vulnerability and load level. The assessment results are as follows: Figure 1 As shown.

[0060] 1. Node vulnerability assessment

[0061] First, establish the system node set and the power source and load sets. Based on the power flow calculation results, statistically analyze the power transfer paths for different power source-load combinations and their power flow components through each node. The power flow betweenness of node i can be defined as the sum of the proportions of power flow passing through node i in the power transfer of all source-load pairs:

[0062]

[0063] In the formula, NPFB(i) is the power flow betweenness of node i. and P represents the index of the power supply set and the load set, respectively. gd (i) represents the power flow component via node i, P gd The total power from the power source to the load, w gd P represents the weight of the source-load combination. g For power output, P d To meet load requirements, the node vulnerability weight w is obtained through normalization. i The higher the weight, the more important the node is in the critical power transfer channel, and the greater the impact of its outage on the system's power supply security and recovery capability.

[0064] 2. Load Classification

[0065] To reflect the varying importance and interruptibility of different electricity services, this embodiment divides the load at each node into four categories proportionally: Category I consists of lifeline and critical loads, such as hospitals, important communications facilities, and emergency centers; Category II consists of essential public services and critical industrial loads; Category III consists of general public services and general industrial loads; and Category IV consists of interruptible or transferable loads, such as some delayed industrial production and commercial loads. A weight π is assigned to each load level. j The following conditions must be met: Class I weight is greater than Class II weight, Class II weight is greater than Class III weight, and Class III weight is greater than Class IV weight.

[0066] 3. Construction of weighted unpowered quantities

[0067] Based on this, the amount of power not supplied at each node and at each level in each time period is defined. The system-weighted power outage can be expressed as a weighted sum of power outages across all nodes, all load levels, and all time periods, with the weights determined by the node vulnerability weights w. i and load level weight π j Together, they can prioritize power supply to highly vulnerable nodes and high-priority loads during the optimization process:

[0068]

[0069] In the formula, This represents the weighted average amount of electricity not supplied in the previous period. For the index of the node set, An index for the load level set. w is the index of the discrete-time set of the day-ahead plan. i π represents the node vulnerability weight. j For load level weighting, To reduce load power, Δt DA This refers to the time step. This metric, as the core objective of resilient scheduling in this invention, is used to prioritize power supply to highly vulnerable nodes and high-priority loads during the optimization process.

[0070] Example 3: Multi-timescale resilient scheduling framework

[0071] like Figure 2 As shown, this embodiment constructs a multi-timescale resilience scheduling framework consisting of day-ahead coordination optimization, intraday rolling scheduling, and real-time resilience-oriented model predictive control. Figure 3 This is a schematic diagram of the multi-timescale resilient scheduling execution process.

[0072] 1. Recent coordination and optimization

[0073] The day-ahead phase uses a 1-hour time resolution to perform static optimization for the entire next day's cycle. A lexicographical optimization method is employed. The first phase aims to minimize the weighted unsupplied power generation, determining the resilience benchmark under extreme high-temperature scenarios. The second phase maximizes the overall system benefits within the allowable weighted unsupplied power generation deviation range, including generation revenue, energy storage operating costs, renewable energy curtailment penalty costs, demand response subsidy costs, and direct load shedding penalty costs. Optimization results include thermal power unit start-up and shutdown plans, pumped storage pumping and generation timing, electrochemical energy storage charge-discharge reference trajectories, and price-based and partially incentive-based demand response schemes. These discrete decision variables and resilience benchmarks are passed to the intraday and real-time phases via memory variables.

[0074] 2. Intraday rolling scheduling

[0075] During the intraday phase, with a time resolution of less than 15 minutes and a rolling outlook window of several hours, the system status and memory variables are read at the beginning of each rolling window. Based on updated wind power and photovoltaic output forecasts and load forecasts, the output of thermal power units, electrochemical energy storage, pumped storage, and intraday stimulus-type demand response are optimized. This phase maintains the unit start-up and shutdown status determined a day before, allowing adjustments to unit output and energy storage charging and discharging power within a continuous decision space. Under the premise of meeting the energy storage state of charge constraints and rolling outage window constraints, operating costs and weighted unsupplied power are minimized as much as possible.

[0076] 3. Real-time resilience-guided model predictive control

[0077] In the real-time phase, a model predictive control method is employed, with time steps of less than 5 minutes and prediction domains of several steps. Within each real-time rolling window, the current system state and ambient temperature are acquired. Ultra-short-term forecasts are used to obtain the net load and renewable energy output trajectories for the next few steps. Using the weighted sum of operating cost and weighted unsupplied power as the objective, and under the conditions of satisfying equipment physical constraints, temperature effect constraints, energy storage state of charge constraints, and rolling outage window constraints, the optimal control sequence within the prediction domain is solved, and only the first control variable is implemented. This process is repeated as time progresses, achieving rolling optimization and feedback correction.

[0078] Example 4: Rolling Stop Window Constraint

[0079] This embodiment achieves fair control over the power supply sequence of loads through a rolling shutdown window constraint. For each load node i, a full power supply indication variable is introduced in each scheduling phase. This variable is a binary variable, when When, the node's unpowered amount during time period t is constrained to be zero; when At that time, a certain percentage of load reduction is allowed during that period. For any continuous sliding time window, the sum of the full power supply indication variables within that window is constrained to be no less than 1, i.e.:

[0080]

[0081] In the formula, and These are the node load reduction and the total node load, respectively. This is a binary variable indicating full power supply; a value of 1 indicates that load shedding is not allowed during that period. T represents the total number of periods in the daily and real-time rolling windows. This constraint ensures that each load node is fully powered for at least one period within any consecutive timeframe, thereby suppressing prolonged continuous power outages. The number of consecutive periods can be set according to grid operation requirements and user capacity, for example, it can be 4 hours or longer.

[0082] Example 5 Typical Application Scenarios

[0083] This embodiment uses a regional wind-solar-hydro-thermal-storage multi-energy complementary system as an example to verify the effectiveness of the method of the present invention. The system includes several thermal power units, a hydropower station, a pumped storage power station, an electrochemical energy storage power station, and large-scale wind and photovoltaic power stations, configured with reserves according to a certain proportion of net load. The typical example system topology and equipment configuration are as follows: Figure 4 As shown, a typical high-temperature day was selected as the research object, and extreme high-temperature scenarios were constructed using actual or synthetic ambient temperature, wind speed, solar irradiance, and load prediction data.

[0084] First, temperature effect models for photovoltaic, thermal power, transmission lines, transformers, and loads are established using the method in Example 1 to obtain the available capacity and temperature-sensitive loads for each time period. Second, vulnerability assessments of system nodes are performed according to the method in Example 2, and the loads of each node are classified into levels, constructing a weighted index of unsupplied power. Then, following the multi-timescale resilient scheduling framework in Example 3, day-ahead coordination optimization, intraday rolling scheduling, and real-time resilience-oriented model predictive control are executed sequentially to obtain the scheduling results for each stage. The system power balance comparison curve is shown below. Figure 5 As shown.

[0085] By comparing scenarios that do not consider temperature effects, do not introduce resilience targets, or only use single-time-scale scheduling, the following typical conclusions can be drawn: Under extreme high-temperature conditions, if the temperature effect is ignored, the output of thermal power units and transmission lines in the scheduling solution will be close to or exceed the actual available capacity, posing a significant operational risk; after introducing the temperature effect, the available capacity and safety margin of the system will decrease significantly, but by coordinating the use of energy storage and demand response through the method of this invention, the load shedding level and the amount of abandoned new energy can be effectively reduced; the multi-time-scale scheduling framework can make full use of intraday and real-time forecast updates, and further compress the weighted unsupplied power and operating costs through rolling optimization and feedback correction; the rolling outage window constraint can significantly reduce long-term continuous power outage events and improve the fairness of power outages among different nodes.

[0086] Example 6: Computer-readable storage medium

[0087] This embodiment provides a computer-readable storage medium, which may be a read-only memory, a disk, or an optical disk, etc., storing a computer program thereon. When executed by a processor, the computer program calls the source-grid-load-storage model, temperature effect parameters, and real-time measurement data of the power system dispatching platform, and executes the multi-timescale resilient dispatching strategy of the multi-energy complementary system under extreme high-temperature conditions according to the steps described in Embodiments 1 to 5. It automatically generates day-ahead, intraday, and real-time dispatching plans, and issues dispatching instructions to the corresponding power generation, energy storage, and load control devices, thereby realizing the engineering application of the method of this invention in a dispatch automation system.

[0088] It should be noted that the specific model parameters, thermal stability coefficient, demand response subsidy price, rolling window length, and time resolution involved in the various embodiments of the present invention can be adjusted according to the operating characteristics and management requirements of different regional power grids, and still fall within the protection scope of the present invention.

[0089] The specific embodiments of the present invention have been described in detail above. However, those skilled in the art should understand that various modifications or variations can be made to the present invention without departing from the spirit and scope of the claims, and all such modifications or variations should be considered to fall within the protection scope of the present invention.

Claims

1. A method for implementing a multi-timescale resilient scheduling strategy for a multi-energy complementary system under extreme high-temperature conditions, characterized in that, Includes the following steps: (1) Temperature effect modeling steps: Based on the ambient temperature and meteorological forecast data of the target area, temperature correction is performed on the operating characteristics of various power sources, networks, and loads in the multi-energy complementary system to establish a unified source-grid-load-storage temperature effect model, including: a. Based on the relationship between ambient temperature and the temperature of photovoltaic power station cells, adjust the upper limit of photovoltaic unit output to obtain the usable photovoltaic capacity that varies with temperature; P PV =P STC ·or std [1-c PV ·(T cell -T ref,PV )] (1) In the formula, P PV P represents the available output of the photovoltaic array under current conditions. STC η is the rated output under standard test conditions. std For the inverter's equivalent efficiency under standard conditions, γ PV T is the temperature power coefficient of a photovoltaic module. cell For the cell temperature, T ref,PV This is the photovoltaic reference temperature. b. Based on the efficiency degradation and capacity derating characteristics of thermal power units under cold-end limited conditions, a piecewise linear relationship between the efficiency and upper limit of output of thermal power units and the change of ambient temperature is established to obtain the time-varying available capacity multiplier of thermal power units. or TH,t =the N -a TH [T amb,t -T thr,TH ] + ,0<η TH,t ≤1 (2) f t =min{φ max ,f max -or thr T amb,t },φ t ≥0 (3) In the formula, η TH,t η is the equivalent power generation efficiency of a thermal power unit. N α is the unit efficiency under rated operating conditions. TH T is the efficiency sensitivity coefficient to ambient temperature. amb,t For ambient temperature, T thr,TH The threshold temperature at which the efficiency of thermal power generation begins to be affected by temperature is [ ]. + For positive part operators, φ t φ is the available capacity multiplier for time period t. max η represents the upper limit of the available capacity of thermal power units. thr The temperature sensitivity coefficient of the thermal power unit; c. Based on the influence of ambient temperature on load level, establish a temperature-sensitive load model, introduce a temperature gain factor to characterize the gain of temperature-controlled load, and map the load increase caused by high temperature into a time-varying gain term of net load. L t =L base +β load [T amb,t -T thr,load ] + (4) In the formula, L t Let L be the load level for time period t. base For the baseline load, β load T is the sensitivity coefficient of the load to temperature. amb,t For ambient temperature, T thr,load The threshold temperature at which temperature-sensitive loads begin to increase; d. Based on the transformer hot spot temperature calculation model, determine the allowable load factor characterizing the transformer capacity derating, and map the ambient temperature to the time-varying capacity upper limit of the transformer. k TR,t =1-β TR (T amb,t -T ref,TR ),0≤κ TR,t ≤1 (5) In the formula, κ TR,t β is the allowable load factor of the transformer during time period t. TR T represents the derating slope of the transformer capacity with respect to ambient temperature. amb,t For ambient temperature, T ref,TR The rated operating ambient temperature of the transformer; e. Based on the thermal balance relationship of transmission line conductors, calculate the upper limit of dynamic thermal stability current of transmission lines under different ambient temperatures, and map the upper limit of dynamic thermal stability current to the upper limit of branch power transmission to form a network transmission capacity constraint that takes into account the impact of extreme high temperature. In the formula, This represents the upper limit of dynamic thermal stability that the line can transmit. β represents the rated transmission capacity under reference conditions. line T is the derating factor for the transmission capacity of a line due to temperature. amb,t For ambient temperature, T ref,line The rated operating temperature of the transmission line; (2) Multi-timescale scheduling framework construction steps: Divide the scheduling time into day-ahead stage, intraday stage and real-time stage, so that each stage shares the source-grid-load-storage temperature effect model and equipment operation constraints, and transmit the unit start-up and shutdown status, energy storage charge status, rolling outage window status and weighted unsupplied power resilience benchmark between each stage through memory variables. (3) Day-ahead scheduling steps: Using hours as the time resolution, under the influence of node vulnerability weights and load level weights, the first-stage objective is to obtain a resilient benchmark solution considering the impact of extreme high temperatures, with the system-weighted unsupplied power as the objective. Under the constraint that the system-weighted unsupplied power is not inferior to the resilient benchmark solution, the second-stage objective is to maximize the comprehensive benefits consisting of power generation revenue, energy storage operating costs, demand response subsidy costs, renewable energy curtailment penalty costs, and direct load shedding penalty costs. In the formula, F DA The system's overall benefits at the current stage are represented by t, where t is the scheduling time variable. and These are the indices for the power source set, renewable energy set, energy storage set, discrete time step of the day-ahead dispatch phase, and load set, respectively. g,t For the net benefit of thermal power units, For renewable energy revenue, For the benefits of hybrid energy storage systems, For the current demand response cost, C shed To reduce load, Δt DA The time step for the day-ahead scheduling; The start-up and shutdown status of thermal power units, pumped storage unit pumping and power generation sequence, electrochemical energy storage charging and discharging plan, and different types of demand response call schemes are optimized, and discrete decision variables and resilience benchmarks are passed to the intraday stage as memory variables; (4) Intraday Rolling Scheduling Steps: With a time resolution of no more than 15 minutes and a rolling outlook window of several hours, the memory variables are read and, based on the updated renewable energy output and load forecast information, the active power output of thermal power units, the charging and discharging power of electrochemical energy storage, the output of pumped storage, and the amount of incentive-based demand response are rolled out for optimization. The objective function is to maximize the overall system benefit during the intraday phase, and its form is as follows: In the formula, F ID (n) represents the overall system benefit within the nth rolling outlook window during the intraday phase, where n is the rolling outlook window number. Let Δt be the set of time steps covered by the nth rolling window during the intraday phase. ID For intraday scheduling time steps, This refers to the cost of responding to demand during the intraday phase. While maintaining the unit start-up and shutdown status and some demand response schemes consistent with the day-ahead phase, constrain the weighted unpowered amount of each node in the current window to not exceed the resilience benchmark of the corresponding time period, and ensure at least one hour of full power supply in any consecutive preset number of rolling outage windows; pass the optimized energy storage charge status, rolling outage window status and weighted unpowered amount of each node as updated memory variables to the next day's rolling window or real-time phase; (5) Real-time resilience-oriented scheduling steps: With a time step of no more than 5 minutes and a prediction domain length of several time steps, update the system state variables and ambient temperature based on real-time measurement data. Use ultra-short-term forecasts to obtain the net load and renewable energy output trajectory for the next few steps. The objective function is to maximize the overall system benefit in the real-time stage, and its form is as follows: In the formula, F RT (n) represents the overall system benefits during the real-time phase. Let Δt be the set of time steps covered by the nth scrolling window in the real-time phase. RT To schedule the time step in real time, Cost of responding to real-time demand; Under the conditions of satisfying the physical constraints of source-grid-load-storage, temperature effect constraints, energy storage state of charge constraints and rolling shutdown window constraints, the control sequence of marginal output of thermal power units, electrochemical energy storage output and incentive demand response call amount in the prediction domain is optimized. Only the first step of the control sequence is implemented, the system state and memory variables are updated and the time is advanced by one scheduling step. The above optimization process is repeated to realize real-time rolling feedback correction of resilient scheduling at multiple time scales. (6) Dispatch instruction generation steps: Based on the multi-timescale dispatch results obtained in the day-ahead, intraday and real-time stages, generate generator output plans, energy storage charging and discharging instructions and demand response call strategies for extreme high temperature scenarios, and send them to the corresponding power supply, energy storage and load-side execution devices to realize multi-timescale resilient dispatch of multi-energy complementary systems under extreme high temperature conditions.

2. The method for implementing the multi-timescale resilient scheduling strategy of a multi-energy complementary system under extreme high-temperature conditions according to claim 1, characterized in that, The construction of the system's weighted unpowered quantity includes node vulnerability, load level, and load loss quantity; Establish a set of nodes and a set of load levels for a multi-energy complementary system, and use the node power flow betweenness index to assess the vulnerability of each node, thereby obtaining the vulnerability weight of each node. In the formula, NPFB(i) is the power flow betweenness of node i. and P represents the index of the power supply set and the load set, respectively. gd (i) represents the power flow component via node i, P gd The total power from the power source to the load, w gd For the source load pair weights, P g For power output, P d For load demand; The total load of each node is divided into at least four levels according to its importance, and a corresponding load level weight is set for each level. The unpowered quantity of each node at each level is weighted and summed according to the vulnerability weight and load level weight to obtain the objective function of the system weighted unpowered quantity. This is to prioritize the power supply to highly vulnerable nodes and high-level loads during the optimization process, as shown below: In the formula, This represents the weighted average amount of electricity not supplied in the previous period. For the index of the node set, An index for the load level set. w is the index of the discrete-time set of the day-ahead plan. i π represents the node vulnerability weight. j For load level weighting, To reduce load power, Δt DA For time step.

3. The method for implementing the multi-timescale resilient scheduling strategy of a multi-energy complementary system under extreme high-temperature conditions according to claim 1 or 2, characterized in that, The rolling supply stop window constraint includes: In the formula, and These are the node load reduction and the total node load, respectively. The binary variable is for full power supply indication. A value of 1 indicates that load shedding is not allowed during this period. T is the total number of periods in the daily and real-time rolling window. A full power supply indicator variable is introduced for each load node. When the full power supply indicator variable is 1, the amount of power outage during the corresponding time period is constrained to be zero. For any consecutive preset number of hours of sliding time window, the sum of the full power supply indicator variables within the window is constrained to be no less than 1, thereby avoiding long-term continuous power outages at the load node and improving the fairness of the power outage sequence.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the method for implementing a multi-timescale resilient scheduling strategy for a multi-energy complementary system under extreme high-temperature conditions as described in any one of claims 1 to 3.

Images