A multi-resource coordinated dispatching system and method for Shaguo Barren New Energy Base

Abstract

The application discloses a kind of multi-resource linkage scheduling systems and methods for Shagehuang new energy base, belong to the technical field of power system dispatching automation.The system includes ecological factor monitoring and output correction module, network type resource partition aggregation and active support module, computing power load cross-regional elastic migration module and multi-resource joint scheduling module.The method comprises: obtaining the ecological factor data of Shagehuang photovoltaic sand control base, establishing the quantitative correlation model of ecological factor and photovoltaic module output attenuation, dynamically correcting the photovoltaic output prediction curve.The application solves the technical problems in the prior art that the photovoltaic sand control ecological factor is not considered, the network type resources are not aggregated, and the computing power load is not involved in the regulation of the sending end power grid, significantly improving the new energy consumption capacity and the safe and stable operation level of the sending end power grid.

Description

Technical Field

[0001] This invention relates to the field of power system dispatch automation technology, specifically to a multi-resource collaborative dispatch system and method for the Shagohuang New Energy Base. Background Technology

[0002] With the deepening of the "dual-carbon" strategy, new energy bases in my country's desert and Gobi regions have entered a stage of large-scale development. Taking Gansu Province as an example, by the end of 2025, the installed capacity of new energy had exceeded 80 million kilowatts, accounting for more than 64%. Ultra-high voltage direct current (UHVDC) transmission channels such as Gansu-Hunan and Gansu-Shandong power transmission have been completed, and the construction of Gansu-Zhejiang and Gansu-Sichuan power transmission projects is being accelerated. In desert and Gobi regions, a photovoltaic desertification control model of generating electricity on the panels and restoring the environment beneath them is commonly adopted. Ecological factors such as vegetation growth beneath the panels and dust accumulation on the modules have a significant impact on photovoltaic output, but the existing dispatch system has not yet incorporated ecological monitoring data into the output prediction and correction process.

[0003] On the grid side, the Shagohuang renewable energy base is transmitted over long distances via UHVDC, forming a weak sending-end grid. The spatiotemporal distribution of grid intensity in various renewable energy gathering zones is uneven, making traditional grid-based control inadequate for low short-circuit ratio conditions. Existing grid-based resource control technologies mostly focus on the adaptive parameters of individual devices, lacking aggregation management and cross-regional support mechanisms for multi-zone grid-based resources, resulting in insufficient stability margin of the sending-end grid. On the load side, the Qingyang "East Data West Computation" national hub node in Gansu Province carries a large-scale computing load. Currently, computing power scheduling and power dispatch are independent, with the computing center only connected to the grid as a rigid load. The spatiotemporal migration potential of its flexible computing tasks has not been explored, failing to provide peak-shaving and frequency regulation support for the sending-end grid. Therefore, we propose a multi-resource collaborative scheduling system and method for the Shagohuang renewable energy base. Summary of the Invention

[0004] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0005] This invention solves the above-mentioned technical problems by adopting the following technical solution: It provides a multi-resource collaborative scheduling system and method for new energy bases in desert areas, comprising the following steps:

[0006] S1: Obtain ecological factor data of the photovoltaic desertification control base in the desert, including vegetation height under the panels, dust density of the modules, surface albedo and soil moisture content; establish a quantitative correlation model between the ecological factor data and the output decay of the photovoltaic modules; based on the quantitative correlation model, dynamically correct the photovoltaic output prediction curve to obtain the corrected photovoltaic output prediction curve.

[0007] S2: Based on the spatiotemporal distribution of the sending-end grid intensity, the grid-type resources are zoned and aggregated; according to the electrical distance between each aggregation zone and the UHVDC converter station, the penetration rate of new energy sources, and the current operating mode, the equivalent short-circuit ratio and inertia level of each zone are calculated; the grid-type energy storage and grid-type SVG dispersed in each aggregation zone are zoned and aggregated according to the grid intensity level, and a two-level architecture of zone aggregation body and dispatch center is established; when the grid intensity of a certain zone is lower than the safety threshold, the grid-type resources of that zone are triggered to switch to the grid control mode. If the support margin is still insufficient, the energy storage of adjacent zones is dispatched to provide cross-regional inertia support through power transfer;

[0008] S3: Obtain cross-regional channel status data between the Hexi Shagohuang Base and the Longdong Computing Center; divide the computing tasks of the computing center into rigid tasks and elastic tasks according to the latency tolerance; based on the corrected photovoltaic output prediction curve and cross-regional channel status data, migrate the elastic computing load across regions to achieve spatiotemporal matching between computing load and new energy output.

[0009] S4: Using the modified photovoltaic output prediction curve as input, and with the grid-type resource support margin and computing load virtual flexibility as constraints, establish a multi-resource joint scheduling model to solve for the solar thermal storage heat release plan, electrochemical energy storage charging and discharging strategy, UHVDC transmission curve and computing load migration plan.

[0010] Preferably, the quantitative correlation model in step S1 is:

[0011] P(t, Correction) = P(t, Prediction) × η(t, Dust Accumulation) × η(t, Vegetation) × η(t, Albedo)

[0012] Wherein, P(t, Correction) is the photovoltaic output after correction at time t, P(t, Prediction) is the original predicted output at time t, η(t, Dust Accumulation) is the dust transmission coefficient, η(t, Vegetation) is the vegetation shading coefficient, and η(t, Albedo) is the albedo correction coefficient.

[0013] The dust transmission coefficient η(t, dust accumulation) = 1 − αd × ρ(t, dust accumulation), where αd is the dust attenuation coefficient and ρ(t, dust accumulation) is the dust density of the component at time t.

[0014] The vegetation shading coefficient η(t, vegetation) = 1 − βv × h(t, vegetation) / H(installation), where βv is the vegetation shading influence coefficient, h(t, vegetation) is the vegetation height under the panel at time t, and H(installation) is the installation height of the photovoltaic module.

[0015] Preferably, the equivalent short-circuit ratio in step S2 is: SCR(i) = Ssc(i) / Pnec(i), where SCR(i) is the equivalent short-circuit ratio of the i-th aggregation zone, Ssc(i) is the short-circuit capacity, and Pnec(i) is the installed capacity of new energy; the inertia level is: Heq(i) = Σ H(j) × S(j) / Sbase, where Heq(i) is the equivalent inertia constant, H(j) is the inertia constant of the j-th grid-type resource, S(j) is the rated capacity, and Sbase is the capacity baseline value.

[0016] Preferably, the safety threshold in step S2 is: when SCR(i) < SCRmin and Heq(i) < Hmin, the power grid strength of the partition is determined to be lower than the safety threshold; SCRmin is the minimum allowable short-circuit ratio and Hmin is the minimum allowable inertia constant.

[0017] Preferably, the determination condition for the cross-regional migration of elastic computing power load in step S3 is as follows: when P(t, wind power) + P(t, correction) − P(t, base load) > Pwaste, the elastic computing power tasks of the Longdong computing power center are migrated to the Hexi Shagohuang base for local consumption; when |Δf(t)| > Δf threshold, virtual inertia support is provided by starting and stopping the elastic computing power load in seconds; P(t, wind power) is the predicted output of wind power, P(t, base load) is the base load, Pwaste is the curtailment power threshold, Δf(t) is the frequency deviation, and Δf threshold is the frequency deviation threshold.

[0018] Preferably, the objective function of the multi-resource joint scheduling model in step S4 is: min C = C(curtailment) + C(energy storage) + C(GFM) + C(computing power), where C(curtailment) is the cost of curtailment of wind and solar power, C(energy storage) is the energy storage operation cost, C(GFM) is the operation cost of grid-type resources, and C(computing power) is the computing power migration cost; the constraints include: power balance constraint P(t, wind power) + P(t, correction) + P(t, CSP) + P(t, ESS) + P(t, GFM) = P(t, DC) + P(t, load) + P(t, computing power); grid strength constraint SCR(i,t) ≥ SCRmin, Heq(i,t) ≥ Hmin; energy storage SOC constraint SOCmin ≤ SOC(t) ≤ SOCmax; computing power delay constraint D(t, computing power) ≤ Dmax.

[0019] Preferably, steps S1 to S4 are executed in a rolling manner across multiple time scales: day-ahead, intraday, and real-time. Step S1 is executed every 15 minutes at the day-ahead scale to generate a corrected day-ahead photovoltaic output prediction curve and formulate a day-ahead solar thermal storage heat release plan and a computing load pre-migration plan. Step S2 is executed every 15 minutes at the intraday scale to update the grid intensity distribution map and adjust the aggregation status of grid-type resource zones. Steps S3 and S4 are executed every 5 minutes at the real-time scale to adjust the energy storage charging and discharging strategy and the second-level response of computing load based on real-time ecological monitoring data and frequency deviation.

[0020] A multi-resource collaborative scheduling system for new energy bases in desert areas includes:

[0021] The ecological factor monitoring and output correction module is used to acquire ecological factor data of the photovoltaic desertification control base in the desert, establish a quantitative correlation model between ecological factors and photovoltaic module output decay, and dynamically correct the photovoltaic output prediction curve.

[0022] The grid-type resource partitioning aggregation and active support module is used to perform partitioning aggregation of grid-type resources based on the spatiotemporal distribution of the sending-end power grid intensity, establish a two-level architecture of partitioning aggregation body and dispatch center, and realize the rolling deployment and decommissioning of grid-type resources and cross-regional support.

[0023] The cross-regional elastic migration module for computing load is used to obtain cross-regional channel status data between the Hexi Shagohuang Base and the Longdong Computing Center, and to migrate the elastic computing load across regions to achieve spatiotemporal matching between computing load and new energy output.

[0024] The multi-resource joint scheduling module is used to establish a multi-resource joint scheduling model by taking the corrected photovoltaic output prediction curve as input and the grid-type resource support margin and computing power load virtual flexibility as constraints, and then solving for the scheduling plan.

[0025] Compared with existing technologies, this invention provides a multi-resource collaborative scheduling system and method for new energy bases in desert areas, which has the following beneficial effects:

[0026] 1. This invention establishes a quantitative correlation model between ecological factors in photovoltaic desertification control and power output attenuation, dynamically embedding ecological monitoring data into the power output prediction and scheduling process, thus solving the technical problem of low power output prediction accuracy in existing photovoltaic desertification control scenarios. Through a grid-based resource zoning aggregation and rolling deployment / retreat mechanism, it achieves zonal autonomy and cross-regional inertia mutual assistance for the weak-sending-end power grid, improving the adaptability of the sending-end power grid to high-proportion renewable energy access. Furthermore, through the flexible cross-regional migration of computing load, the "east-to-west computing" computing center is transformed into a virtual flexible resource, expanding the space for local renewable energy consumption. The coordinated efforts of these three aspects significantly improve the renewable energy utilization rate of the desert and Gobi renewable energy base and the safe and stable operation level of the sending-end power grid.

[0027] 2. Specifically, dynamic correction of ecological factors reduces photovoltaic output prediction errors, providing a data foundation for the accurate formulation of thermal energy storage, energy storage, and power transmission curves, avoiding power curtailment or grid impact caused by prediction deviations. Active support for grid-based resource zoning and aggregation provides a safety boundary for thermal energy storage heat release and energy storage charging and discharging, allowing physical energy storage to perform power regulation over a wider range without triggering grid instability, while ensuring power quality during computing load migration. The equivalent virtual inertia and peak-shaving capabilities provided by cross-regional computing load migration reduce the charging and discharging depth and frequency of physical energy storage, delaying the capacity decay of electrochemical energy storage and reducing thermal stress fatigue in solar thermal energy storage systems. These three factors form a positive feedback loop of "precise prediction—active support—flexible load": accurate ecological correction prediction provides a time window for pre-adjustment of grid-based resources, grid-based support creates grid conditions for the safe migration of computing loads, and the flexible response of computing loads, in turn, reduces the pressure on physical energy storage, enabling thermal and energy storage systems to perform power balancing tasks more efficiently. This linkage mechanism is particularly suitable for scenarios where large-scale new energy bases and computing centers in desert and Gobi areas such as Gansu are spatially separated, providing technical support for the integrated development of new energy, desertification control and computing power. Attached Figure Description

[0028] Figure 1 This is a diagram of the multi-resource linkage scheduling system architecture for the Shagohuang New Energy Base of this invention.

[0029] Figure 2 This is a flowchart of the multi-time-scale rolling scheduling method of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] This embodiment uses the Gansu Hexi Desert New Energy Base as the application object. This base includes the Jiuquan 10-million-kilowatt wind power base, the Zhangye 1-million-kilowatt photovoltaic power base, and supporting solar thermal power plants. Power is transmitted to Hunan via the Gansu-to-Hunan ultra-high-voltage direct current transmission project. The sending-end power grid includes five new energy aggregation zones. The Qingyang "East Data West Computing" computing center is located in the eastern Gansu region and is interconnected with the Hexi base through a cross-regional communication channel.

[0032] like Figure 1As shown, the system of this invention includes an ecological factor monitoring and output correction module, a network-type resource partitioning aggregation and active support module, a computing load cross-regional elastic migration module, and a multi-resource joint scheduling module. The ecological factor monitoring and output correction module collects data such as under-panel vegetation height and component dust density via drones / robots; the network-type resource partitioning aggregation and active support module connects network-type energy storage and network-type SVG in each partition; the computing load cross-regional elastic migration module is interconnected with the Qingyang computing center through a cross-regional channel; and the multi-resource joint scheduling module outputs a solar thermal storage heat release plan, an energy storage charging and discharging strategy, an ultra-high voltage DC transmission curve, and a computing load migration plan.

[0033] like Figure 2 As shown, the method of the present invention is executed on a rolling basis according to multiple time scales, including day-ahead, intraday, and real-time.

[0034] Day-ahead scale (T-1 day):

[0035] Step S1: The ecological factor monitoring and output correction module acquires ecological factor data for each photovoltaic power station. Taking a 100MW photovoltaic desertification control power station in Zhangye as an example, the drone inspection found that the dust density of the modules the next day was ρ(t, dust accumulation) = 1.2g / m², the average height of vegetation under the panels was h(t, vegetation) = 0.8m (module installation height H(installation) = 2.5m), and the surface albedo changed by +5% compared to the baseline value. Substituting into the quantitative correlation model:

[0036] η(t, dust accumulation) = 1 − 0.015 × 1.2 = 0.982

[0037] η(t, vegetation) = 1 − 0.12 × 0.8 / 2.5 = 0.9616

[0038] η(t, albedo) = 1 + 0.03 × 0.05 = 1.0015

[0039] The product of the correction factors is 0.982 × 0.9616 × 1.0015 ≈ 0.945. If the original predicted output P(t, prediction) = 80MW, then the corrected output P(t, correction) = 75.6MW.

[0040] Step S2: The grid-type resource partition aggregation and active support module calculates the grid strength of each partition. Taking partition 1 as an example, the short-circuit capacity Ssc(1) = 1500MVA, the new energy installed capacity Pnec(1) = 1200MW, then SCR(1) = 1.25; the grid-type energy storage inertia constant H(j) = 5s, the rated capacity S(j) = 200MVA, then Heq(1) = 5 × 200 / 1000 = 1.0s. Set SCRmin = 1.2, Hmin = 0.8s, this partition is at the safety boundary.

[0041] Step S3: The cross-regional elastic migration module for computing power load obtains the predicted total output of new energy at the Hexi base for the next day: P(t, wind power) + P(t, correction) = 3500MW, and the base load P(t, base load) = 1800MW. A pre-migration plan for elastic computing power is formulated: approximately 200MW of elastic load from the Qingyang computing center is pre-migrated to the Hexi base to match the midday photovoltaic peak.

[0042] Step S4: The multi-resource joint scheduling module establishes a day-ahead joint scheduling model. The cost coefficients in the objective function min C are: curtailment penalty 50 yuan / MWh, energy storage operation cost 30 yuan / MWh, grid-type resource operation cost 80 yuan / MWh, and computing power migration cost 20 yuan / MWh. The solution yields: the next-day heat release power generation plan for the solar thermal and thermal storage power plant is 200MW during peak hours to smooth out photovoltaic fluctuations; the electrochemical energy storage charging and discharging strategy is charging during off-peak hours and discharging during peak hours; and the UHVDC transmission curve is a stable transmission of 8000MW±5%.

[0043] Intraday scale (every 15 minutes): Update ecological factor data and recalculate the grid intensity distribution map. If the penetration rate of new energy in zone 3 increases due to a sudden increase in wind speed, and the SCR(3) drops to 1.15, the 100MW grid-type energy storage in this zone will be switched to grid control mode, providing virtual inertia ΔH=0.5s. If the support margin is still insufficient, the 50MW energy storage in zone 2 will be dispatched to provide cross-regional inertia support through power transfer.

[0044] Real-time scale (every 5 minutes): When the frequency deviation of the sending-end power grid Δf(t) = +0.15Hz exceeds the threshold Δf threshold = 0.1Hz, the computing power load cross-regional elastic migration module starts the local 200MW elastic computing power load of the Hexi base in seconds to absorb the surplus new energy power; at the same time, the multi-resource joint scheduling module reduces the energy storage charging power by 50MW, and the solar thermal energy storage system increases the thermal storage power by 30MW, and works together to smooth out frequency fluctuations.

[0045] Unlike the above embodiments, ecological factor data can also be obtained through the fusion of satellite remote sensing and IoT sensors. The vegetation height data is updated every 7 days, and the dust accumulation density of the photovoltaic modules is monitored in real time by a light intensity sensor on the surface of the photovoltaic modules, with an update cycle of 1 hour. When the dust accumulation density exceeds 2.0 g / m² and no effective precipitation is predicted in the next 4 hours, the operation and maintenance linkage unit automatically triggers the cleaning robot's operation command and simultaneously adjusts the energy storage pre-charge and discharge plan: pre-charging to SOC=90% 1 hour before cleaning to cope with the decrease in output during cleaning.

[0046] Table 1. Main parameters of the multi-resource joint scheduling model

[0047] Parameter name symbol Value unit Dust accumulation attenuation coefficient αd 0.015 m² / g Vegetation shadow influence coefficient βv 0.12 — Minimum allowable short-circuit ratio SCRmin 1.2 — Minimum allowable inertia constant Hmin 0.8 s Curtailment power threshold Pwaste 500 MW Frequency deviation threshold Δf threshold 0.1 Hz Energy storage SOC lower limit SOCmin 0.1 — Energy storage SOC limit SOCmax 0.9 — Maximum computing latency Dmax 100 ms

[0048] Through the aforementioned system and method, this invention realizes multi-resource linkage scheduling of ecological factor monitoring, active support of grid-type resources, and flexible migration of computing load in the Shagohuang new energy base, effectively improving the new energy consumption capacity and the safe and stable operation level of the sending-end power grid.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-resource collaborative scheduling method for new energy bases in desert and Gobi areas, characterized in that, Includes the following steps: S1: Obtain ecological factor data of the photovoltaic desertification control base in the desert, including vegetation height under the panels, dust density of the modules, surface albedo and soil moisture content; establish a quantitative correlation model between the ecological factor data and the output decay of the photovoltaic modules; based on the quantitative correlation model, dynamically correct the photovoltaic output prediction curve to obtain the corrected photovoltaic output prediction curve. S2: Based on the spatiotemporal distribution of the sending-end grid intensity, the grid-type resources are zoned and aggregated; according to the electrical distance between each aggregation zone and the UHVDC converter station, the penetration rate of new energy sources, and the current operating mode, the equivalent short-circuit ratio and inertia level of each zone are calculated; the grid-type energy storage and grid-type SVG dispersed in each aggregation zone are zoned and aggregated according to the grid intensity level, and a two-level architecture of zone aggregation body and dispatch center is established; when the grid intensity of a certain zone is lower than the safety threshold, the grid-type resources of that zone are triggered to switch to the grid control mode. If the support margin is still insufficient, the energy storage of adjacent zones is dispatched to provide cross-regional inertia support through power transfer; S3: Obtain cross-regional channel status data between the Hexi Shagohuang Base and the Longdong Computing Center; divide the computing tasks of the computing center into rigid tasks and elastic tasks according to the latency tolerance; based on the corrected photovoltaic output prediction curve and cross-regional channel status data, migrate the elastic computing load across regions to achieve spatiotemporal matching between computing load and new energy output. S4: Using the modified photovoltaic output prediction curve as input, and with the grid-type resource support margin and computing load virtual flexibility as constraints, establish a multi-resource joint scheduling model to solve for the solar thermal storage heat release plan, electrochemical energy storage charging and discharging strategy, UHVDC transmission curve and computing load migration plan.

2. The multi-resource collaborative scheduling method for new energy bases in desert areas according to claim 1, characterized in that, The quantitative correlation model mentioned in step S1 is: P(t, Correction) = P(t, Prediction) × η(t, Dust Accumulation) × η(t, Vegetation) × η(t, Albedo) Wherein, P(t, Correction) is the photovoltaic output after correction at time t, P(t, Prediction) is the original predicted output at time t, η(t, Dust Accumulation) is the dust transmission coefficient, η(t, Vegetation) is the vegetation shading coefficient, and η(t, Albedo) is the albedo correction coefficient. The dust transmission coefficient η(t, dust accumulation) = 1 − αd × ρ(t, dust accumulation), where αd is the dust attenuation coefficient and ρ(t, dust accumulation) is the dust density of the component at time t. The vegetation shading coefficient η(t, vegetation) = 1 − βv × h(t, vegetation) / H(installation), where βv is the vegetation shading influence coefficient, h(t, vegetation) is the vegetation height under the panel at time t, and H(installation) is the installation height of the photovoltaic module.

3. The multi-resource collaborative scheduling method for new energy bases in desert areas according to claim 1, characterized in that, The equivalent short-circuit ratio mentioned in step S2 is: SCR(i) = Ssc(i) / Pnec(i), where SCR(i) is the equivalent short-circuit ratio of the i-th aggregation zone, Ssc(i) is the short-circuit capacity, and Pnec(i) is the installed capacity of new energy; the inertia level is: Heq(i) = Σ H(j) × S(j) / Sbase, where Heq(i) is the equivalent inertia constant, H(j) is the inertia constant of the j-th grid-type resource, S(j) is the rated capacity, and Sbase is the capacity baseline value.

4. The multi-resource collaborative scheduling method for new energy bases in desert areas according to claim 1, characterized in that, The safety threshold mentioned in step S2 is: when SCR(i) < SCRmin and Heq(i) < Hmin, it is determined that the grid strength of the partition is lower than the safety threshold; SCRmin is the minimum allowable short-circuit ratio and Hmin is the minimum allowable inertia constant.

5. A multi-resource collaborative scheduling method for new energy bases in desert areas according to claim 1, characterized in that, The determination condition for the cross-regional migration of elastic computing power load in step S3 is as follows: when P(t, wind power) + P(t, correction) − P(t, base load) > Pwaste, the elastic computing power tasks of the Longdong computing power center are migrated to the Hexi Shagohuang base for local consumption; when |Δf(t)| > Δf threshold, virtual inertia support is provided by starting and stopping the elastic computing power load in seconds; P(t, wind power) is the predicted output of wind power, P(t, base load) is the base load, Pwaste is the curtailment power threshold, Δf(t) is the frequency deviation, and Δf threshold is the frequency deviation threshold.

6. A multi-resource collaborative scheduling method for new energy bases in desert areas according to claim 1, characterized in that, The objective function of the multi-resource joint scheduling model in step S4 is: min C = C(curtailment) + C(energy storage) + C(GFM) + C(computing power), where C(curtailment) is the penalty cost for curtailing wind and solar power, C(energy storage) is the energy storage operation cost, C(GFM) is the grid-type resource operation cost, and C(computing power) is the computing power migration cost. The constraints include: power balance constraint P(t, wind power) + P(t, correction) + P(t, CSP) + P(t, ESS) + P(t, GFM) = P(t, DC) + P(t, load) + P(t, computing power); grid strength constraint SCR(i,t) ≥ SCRmin, Heq(i,t) ≥ Hmin; energy storage SOC constraint SOCmin ≤ SOC(t) ≤ SOCmax; computing power delay constraint D(t, computing power) ≤ Dmax.

7. A multi-resource collaborative scheduling method for new energy bases in desert areas according to claim 1, characterized in that, Steps S1 to S4 are executed in a rolling manner according to multiple time scales including day-ahead, intraday, and real-time: Step S1 is executed at the day-ahead scale to generate the corrected day-ahead photovoltaic power output prediction curve and to formulate the day-ahead solar thermal storage heat release plan and computing load pre-migration plan. Every 15 minutes on an intraday scale, step S2 is executed to update the power grid intensity distribution map and adjust the aggregation status of grid-type resource partitions. Steps S3 and S4 are executed every 5 minutes in real time, adjusting the energy storage charging and discharging strategy and computing load response in seconds based on real-time ecological monitoring data and frequency deviation.

8. A multi-resource collaborative scheduling system for new energy bases in desert and Gobi areas, characterized in that: include: The ecological factor monitoring and output correction module is used to acquire ecological factor data of the photovoltaic desertification control base in the desert, establish a quantitative correlation model between ecological factors and photovoltaic module output decay, and dynamically correct the photovoltaic output prediction curve. The grid-type resource partitioning aggregation and active support module is used to perform partitioning aggregation of grid-type resources based on the spatiotemporal distribution of the sending-end power grid intensity, establish a two-level architecture of partitioning aggregation body and dispatch center, and realize the rolling deployment and decommissioning of grid-type resources and cross-regional support. The cross-regional elastic migration module for computing load is used to obtain cross-regional channel status data between the Hexi Shagohuang Base and the Longdong Computing Center, and to migrate the elastic computing load across regions to achieve spatiotemporal matching between computing load and new energy output. The multi-resource joint scheduling module is used to establish a multi-resource joint scheduling model by taking the corrected photovoltaic output prediction curve as input and the grid-type resource support margin and computing power load virtual flexibility as constraints, and then solving for the scheduling plan.

Images