A method, system, device and medium for airport energy balance management

Abstract

A method, system, device and medium for airport energy balance management, relating to the field of energy management. In this method, an energy topology map is constructed based on multi-source data; the energy load demand of the airport in different time periods is predicted based on the energy topology map, the supply of renewable energy in different time periods is determined based on the predicted light intensity and wind power, and the probability of time-sharing energy supply and demand balance is determined based on the energy load demand and supply; a preliminary energy management plan is determined based on cost parameters, time-sharing energy supply and demand balance probability and equipment operation constraints, the preliminary energy management plan includes the charging and discharging sequence of the energy storage system, the start and stop plan of the backup generator and the proportion of power purchased by the power grid; the preliminary energy management plan is executed through the digital twin of the airport to obtain the first energy execution data, and the preliminary energy management plan is adjusted based on the difference between the first energy execution data and the second energy execution data at the current moment. Implementing the technical solution provided by this application reduces energy costs.

Description

Technical Field

[0001] The present application relates to the technical field of energy management, and in particular to a method, system, device and medium for airport energy balance management. Background Art

[0002] In today's world, airports, as important transportation hubs, are crucial for ensuring efficient and sustainable airport operations. With the rapid development of the air transport industry, airports are expanding in size and the number of flights is increasing, leading to a sharp increase in energy consumption. Furthermore, energy issues have become a global challenge, and energy conservation, emission reduction, and improved energy efficiency are inevitable trends across all industries. As major energy consumers, airports must ensure that energy is rationally allocated and effectively managed, not only reducing operating costs but also minimizing environmental impact and enhancing their overall competitiveness.

[0003] Currently, traditional airport energy management relies primarily on fixed energy supply plans based on experience. Some airports set daily or weekly energy supply quotas based on historical average energy consumption, without considering dynamic changes in actual operations. Other airports have installed simple energy monitoring equipment, but these devices only record total energy usage and cannot achieve refined management.

[0004] These existing technologies have significant flaws. Due to a lack of comprehensive analysis of multi-source data, such as flight dynamics and weather changes, traditional energy supply plans cannot accurately reflect the airport's actual energy needs at different times. Energy usage in various areas of an airport is complex and volatile, and a single monitoring approach cannot promptly identify energy waste or insufficient supply, easily leading to irrational energy allocation. Summary of the Invention

[0005] This application provides a method, system, equipment and medium for airport energy balance management, which accurately predicts airport energy load demand and renewable energy supply, reasonably formulates energy management plans and adjusts them in real time, realizes airport energy balance management, and reduces energy costs.

[0006] In a first aspect of the present application, a method for airport energy balance management is provided, which is applied to an energy management platform. The method comprises:

[0007] Acquiring multi-source data, including flight data, weather data, energy usage data for various areas within the airport, and energy price data, and constructing an energy topology map based on the multi-source data;

[0008] Predicting the energy load demand of the airport at different time periods based on the energy topology map, predicting the light intensity and wind speed based on the meteorological data, determining the supply of renewable energy at different time periods based on the light intensity and wind speed, and determining the time-sharing energy supply and demand balance probability based on the energy load demand and the supply;

[0009] Determining cost parameters based on the energy price data, and determining a preliminary energy management plan based on the cost parameters, the time-sharing energy supply and demand balance probability, and equipment operation constraints, the preliminary energy management plan including the energy storage system charge and discharge sequence, the backup generator start and stop plan, and the power purchase ratio of the power grid;

[0010] The preliminary energy management plan is executed through the digital twin of the airport to obtain first energy execution data, second energy execution data at the current moment is obtained, and the preliminary energy management plan is adjusted according to the difference between the first energy execution data and the second energy execution data.

[0011] Optionally, constructing an energy topology map based on the multi-source data includes:

[0012] Based on the airport's building information model and the coordinates of the physical equipment, the physical equipment is converted into graph nodes with spatial attributes, and the real-time sensor data stream is bound to establish a mapping relationship between the physical equipment and the data;

[0013] According to the energy transmission path, the directed connection edges between the physical devices are constructed, and the hierarchical relationship and capacity constraints are defined to form an initial static topology skeleton;

[0014] The state changes of the physical devices are recorded in a time series database, and an energy topology map with a time stamp is generated according to the state changes and the initial static topology skeleton.

[0015] Optionally, predicting the energy load demand of the airport at different time periods according to the energy topology map includes:

[0016] Extracting node dynamic parameters and device association weights from the energy topology map, and building a multidimensional feature matrix based on historical load patterns;

[0017] Inputting the equipment status data, flight dynamics, meteorological parameters and the multidimensional feature matrix in the energy topology map into a long short-term memory network to generate a benchmark forecast curve for load demand by time period;

[0018] Based on the failure probability of graph nodes, meteorological volatility and the benchmark prediction curve, the Monte Carlo method is used to simulate the preset abnormal scenarios and output the energy load demand in different time periods.

[0019] Optionally, the predicting of light intensity and wind speed according to the meteorological data, and determining the supply of renewable energy in different time periods according to the light intensity and wind speed includes:

[0020] Based on satellite cloud images and ground meteorological station data, the temporal fluctuation characteristics of light intensity and the spatial distribution of wind fields in the future preset time period are extracted to construct a spatiotemporal grid meteorological parameter matrix;

[0021] Predicting the probability distribution of light intensity based on the spatiotemporal gridded meteorological parameter matrix and generating a joint probability density function of renewable energy supply by combining wind turbine power curves and turbulence intensity coefficients;

[0022] The time-division output power of renewable energy is obtained according to the equipment parameters in the energy topology map and the joint probability density function.

[0023] Optionally, determining the time-sharing energy supply and demand balance probability according to the energy load demand and the supply includes:

[0024] Inputting the energy load demand and the supply into a Bayesian network, constructing a supply-demand joint probability model, and generating a time-divided supply-demand gap probability density function;

[0025] Simulating the preset scenario based on the Monte Carlo method, and calculating the expected value and probability of the supply and demand gap in each time period according to the supply and demand gap probability density function of the time period;

[0026] The probabilities are mapped to the digital twin of the airport, and a visual heat map is generated based on operational constraints to identify time periods and areas where the intervention priority is greater than a threshold.

[0027] Optionally, determining a preliminary energy management plan according to the cost parameter, the time-sharing energy supply and demand balance probability, and equipment operation constraints includes:

[0028] Based on the energy price data, the unit energy cost of different energy types in each time period is calculated, and based on the unit energy cost, the charging and discharging cost of the energy storage system, and the start-up and shutdown cost of the backup generator, an energy cost parameter matrix for each time period is constructed;

[0029] Taking minimizing total energy cost as the objective function, a multi-objective optimization model is constructed according to the energy cost parameter matrix, the time-sharing energy supply and demand balance probability and equipment operation constraints;

[0030] The multi-objective optimization model is solved to obtain the energy storage system charge and discharge timing, the backup generator start and stop plan, and the power purchase ratio of the power grid to form a preliminary energy management plan.

[0031] Optionally, adjusting the preliminary energy management plan based on the difference between the first energy performance data and the second energy performance data includes:

[0032] Comparing and analyzing the first energy execution data with the second energy execution data, calculating differences between the first energy execution data and the second energy execution data in terms of operating parameters of each physical device, energy usage parameters of each region, and energy cost parameters, and determining a degree of difference based on the differences;

[0033] determining a target factor causing the difference according to the degree of the difference and the amount of change in the multi-source data of the airport;

[0034] The preliminary energy management plan is adjusted according to the target factors.

[0035] In a second aspect of the present application, a system for airport energy balance management is provided, comprising a map module, a prediction module, a planning module, and an adjustment module, wherein:

[0036] a mapping module configured to acquire multi-source data and construct an energy topology map based on the multi-source data, wherein the multi-source data includes flight data, weather data, energy usage data of various areas within the airport, and energy price data;

[0037] a prediction module configured to predict the energy load demand of the airport at different time periods based on the energy topology map, predict the light intensity and wind speed based on the meteorological data, determine the supply of renewable energy at different time periods based on the light intensity and wind speed, and determine the time-sharing energy supply and demand balance probability based on the energy load demand and the supply;

[0038] a planning module configured to determine cost parameters based on the energy price data, and determine a preliminary energy management plan based on the cost parameters, the time-sharing energy supply and demand balance probability, and equipment operation constraints, the preliminary energy management plan including a charge and discharge sequence of the energy storage system, a start and stop plan for the backup generator, and a proportion of power purchased from the power grid;

[0039] An adjustment module is configured to execute the preliminary energy management plan through the digital twin of the airport to obtain first energy execution data, obtain second energy execution data at the current moment, and adjust the preliminary energy management plan according to the difference between the first energy execution data and the second energy execution data.

[0040] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device performs any of the methods described above.

[0041] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores instructions. When the instructions are executed, any one of the methods described above is executed.

[0042] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0043] 1. By building an energy topology map that integrates flight, weather, and regional energy consumption data, and combining it with long-short-term memory (LSTM) networks and Monte Carlo simulation, we can achieve accurate forecasts of energy load demand and renewable energy supply by time period, effectively reducing supply and demand forecast errors. We also build a joint supply and demand probability model based on a Bayesian network to quantify the probability of supply and demand gaps in each time period. Combined with operational constraints, we generate a visual heat map to achieve a dynamic balance between energy supply and demand, avoiding operational disruptions caused by energy waste or shortages.

[0044] 2. A time-of-use cost function model is established using energy price data. Combined with the probability of supply and demand equilibrium and equipment operating constraints, mixed integer linear programming is used to generate an optimal energy management plan. This improves the charging and discharging efficiency of the energy storage system, reduces the number of backup generator starts and stops, and lowers the grid's electricity purchase costs. The energy storage system's charging and discharging sequence and the grid's electricity purchase ratio are adjusted in real time based on peak and valley electricity prices. For example, energy storage is charged during low-price periods and discharged during peak periods, further reducing energy procurement costs.

[0045] 3. By simulating a preliminary energy management plan using the airport digital twin and comparing the primary energy execution data (simulation results) with the secondary energy execution data (actual data), abnormal events such as equipment failures and sudden weather changes can be quickly identified, automatically triggering a plan adjustment mechanism. Based on a Monte Carlo simulation-based abnormal scenario library and real-time data difference analysis, emergency adjustment plans can be generated to ensure the continued stable operation of the airport energy system.

[0046] 4. Combining satellite cloud images and ground-based meteorological station data to construct a spatiotemporal gridded meteorological parameter matrix, the system quantifies the volatility of renewable energy supply through a joint probability density function, thereby increasing the absorption rate of renewable energy sources such as photovoltaic and wind power. The energy storage system adjusts its charging and discharging strategies in real time based on fluctuations in renewable energy supply. For example, it prioritizes storing electricity when sunlight intensity is high and releasing it when wind power is insufficient, thereby improving energy self-sufficiency.

[0047] 5. The supply-demand gap heat map and intervention priority identification provide operators with intuitive decision-making basis, reduce manual intervention errors, and shorten energy scheduling response time. Through the data closed loop between the digital twin and the actual system, iterative optimization of the energy management plan can be achieved (for example, automatic update of model parameters once a week) to adapt to the dynamic changes in the scale of airport operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flow chart of a method for airport energy balance management disclosed in an embodiment of the present application;

[0049] Figure 2 This is a module diagram of a system for airport energy balance management disclosed in an embodiment of the present application;

[0050] Figure 3 This is a structural diagram of an electronic device disclosed in an embodiment of the present application.

[0051] Explanation of the accompanying drawings: 201, graph module; 202, prediction module; 203, planning module; 204, adjustment module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0053] In the description of the embodiments of this application, words such as "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "for example" or "for instance" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "for example" or "for instance" is intended to present the relevant concepts in a concrete manner.

[0054] In the description of the embodiments of the present application, the term "multiple" means two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0055] This embodiment discloses a method for airport energy balance management, which is applied to the energy management platform. Figure 1 This is a flow chart of a method for airport energy balance management disclosed in an embodiment of the present application, such as Figure 1 As shown, the method includes the following steps:

[0056] S101. Acquire multi-source data and construct an energy topology map based on the multi-source data, wherein the multi-source data includes flight data, weather data, energy usage data of various areas within the airport, and energy price data;

[0057] S102. Predicting the energy load demand of the airport at different time periods based on the energy topology map, predicting the light intensity and wind speed based on the meteorological data, determining the supply of renewable energy at different time periods based on the light intensity and wind speed, and determining a time-sharing energy supply and demand balance probability based on the energy load demand and the supply;

[0058] S103. Determine cost parameters based on the energy price data, and determine a preliminary energy management plan based on the cost parameters, the time-sharing energy supply and demand balance probability, and equipment operation constraints. The preliminary energy management plan includes a charge and discharge sequence for the energy storage system, a start and stop plan for the backup generator, and a proportion of power purchased from the power grid.

[0059] S104. Execute the preliminary energy management plan through the digital twin of the airport to obtain first energy execution data, obtain second energy execution data at the current moment, and adjust the preliminary energy management plan based on the difference between the first energy execution data and the second energy execution data.

[0060] Flight data: This system obtains information related to flight schedules (takeoff and landing times, aircraft types, and passenger capacity), actual operating status (delays / cancellations), and energy consumption of ground support equipment (boarding bridges and jet bridge lighting) through the airport operations system or the Air Traffic Control Bureau interface. Meteorological data: This system integrates satellite cloud images (light intensity forecasts) and ground-based meteorological stations (wind speed and temperature) to construct a spatiotemporal gridded meteorological database. Regional energy consumption data: This system collects real-time electricity and heat consumption data for areas such as terminals, aprons, and runways through smart meters and sensor networks, and categorizes and compiles statistics by functional area (e.g., lighting, air conditioning, and power systems). Energy price data: This system collects real-time grid electricity prices (peak, off-peak, and time-of-use electricity prices), backup generator fuel costs (diesel / natural gas unit price), and energy storage system charging and discharging costs (including battery loss reduction). Based on the airport BIM (Building Information Model) and physical equipment coordinates, transformers, energy storage cabinets, photovoltaic panels, and other components are converted into graph nodes with spatial attributes and bound to real-time sensor data streams (e.g., current, voltage, and power). Directed connections are constructed based on energy transmission paths (such as cable connections and thermal pipelines), defining hierarchical relationships (such as high voltage → medium voltage → low voltage distribution) and capacity constraints (such as maximum cable current carrying capacity) to form an initial static topology skeleton. A time series database records device state changes (such as switch status and load rate) to generate a timestamped dynamic topology graph. Node dynamic parameters (such as real-time device power and historical load curves) and device association weights (such as the correlation between terminal lighting and flight density) are extracted from the energy topology graph. Combined with historical load patterns, a multidimensional feature matrix is ​​constructed. This feature matrix, along with flight dynamics (takeoff and landing times) and meteorological parameters (temperature and humidity), is input into a long-short-term memory network to generate a baseline forecast curve for load demand by time period (such as hourly load values ​​for the next 24 hours). Based on the graph node failure probability (such as historical transformer failure rates) and meteorological volatility (such as wind speed standard deviation), a Monte Carlo method is used to simulate pre-defined abnormal scenarios (such as equipment failures and extreme weather conditions) to output load demand distributions for different time periods. Based on satellite cloud images and ground-based meteorological station data, the temporal fluctuation characteristics of sunlight intensity (such as hourly irradiance) and the spatial distribution of wind fields (such as wind speed gradients in the apron area) for the next 72 hours were extracted to construct a spatiotemporal gridded meteorological parameter matrix. The photovoltaic panel power curves and wind turbine power curves were combined to generate a joint probability density function for renewable energy supply, and the expected value and variance of output power by time period were calculated. Energy load demand and renewable energy supply were input into a Bayesian network to construct a joint supply and demand probability model, generating a time-based supply and demand gap probability density function (such as the hourly gap probability distribution). Based on simulated scenarios, the expected value and probability of the supply and demand gap were calculated for each time period and mapped to the airport digital twin. A visual heat map was generated (red indicates high-risk periods, yellow indicates medium-risk periods, and green indicates low-risk periods). Time periods and areas (such as Terminal 2) with intervention priorities exceeding a threshold (e.g., 80%) were identified.Energy price data is extracted to identify unit prices for grid electricity purchases, backup generator fuel, and energy storage charging and discharging costs at each time period. This is combined with fixed costs (equipment depreciation and maintenance amortization) and variable costs (energy loss) to construct a time-of-use cost function model. A mixed integer linear programming (MILP) model is constructed to minimize total energy costs. Using Gurobi or CPLEX solvers, a preliminary energy management plan is generated, specifying the energy storage system charging and discharging schedule (accurate to the minute), the backup generator start-up and shutdown schedule (including start and stop times and operating durations), and the grid power purchase ratio (accurate to a percentage). This preliminary energy management plan is loaded as input into the airport digital twin, simultaneously initializing the parameters of the physical device models, the energy network model, and the energy flow model. Operations such as energy storage charging and discharging, generator start-up and shutdown, and grid power purchases are simulated on a realistic timescale, updating equipment status and regional energy consumption in real time. Virtual sensors within the digital twin record primary energy execution data (such as actual energy storage charging and discharging power, actual generator output power, and actual grid power purchase). Through a sensor network deployed in the airport's actual energy system (such as smart meters and current transformers), current secondary energy performance data (including actual equipment operating parameters, regional energy consumption, and energy costs) is obtained. The difference between the primary and secondary energy performance data in various parameters (such as energy storage charge and discharge power) is calculated. Based on this difference analysis and the airport's actual operational logs (such as flight dynamics and equipment failure records), combined with the digital twin simulation logic, the main causes of the discrepancy are diagnosed (such as load forecast deviations due to flight delays or abnormal output power due to sudden equipment failures). Based on the diagnostic results, an emergency adjustment plan is quickly generated. The energy storage charge and discharge sequence and the proportion of power purchased from the grid are re-optimized based on actual flight takeoff and landing times and passenger volume. Renewable energy supply forecasts are adjusted based on real-time meteorological data (such as sudden drops in sunlight intensity due to sudden rainfall) to calibrate the probability of energy supply and demand balance. Failure impacts (such as a transformer failure causing load shifting) are simulated in the digital twin to generate backup generator start and stop plans and energy storage coordination strategies. The adjusted energy management plan is re-entered into the digital twin for simulation verification. If verification passes, it is applied to the actual system; otherwise, iterative optimization continues.

[0061] Optionally, constructing an energy topology map based on the multi-source data includes:

[0062] Based on the airport's building information model and the coordinates of the physical equipment, the physical equipment is converted into graph nodes with spatial attributes, and the real-time sensor data stream is bound to establish a mapping relationship between the physical equipment and the data;

[0063] According to the energy transmission path, the directed connection edges between the physical devices are constructed, and the hierarchical relationship and capacity constraints are defined to form an initial static topology skeleton;

[0064] The state changes of the physical devices are recorded in a time series database, and an energy topology map with a time stamp is generated according to the state changes and the initial static topology skeleton.

[0065] Based on the airport's BIM (Building Information Model), the spatial coordinates and structural information (walls, pipes, and cable paths) of each building area (terminal, apron, and energy station) are extracted. Through on-site surveys or IoT device deployment, the physical coordinates (e.g., latitude and longitude, floor, and bay number) of equipment such as transformers, energy storage cabinets, photovoltaic panels, and air conditioning units are obtained and spatially aligned with the corresponding areas in the BIM (with an accuracy of <0.5 meters). The physical device coordinates are mapped to the virtual space of the BIM model, generating graph nodes with spatial attributes (e.g., node coordinates (x, y, z), where z represents the altitude level). Each graph node is associated with a unique device ID and its sensor data (e.g., the node "Transformer T1" is associated with the real-time value of current sensor ID "S001"). Sensor data is processed in real time, and the dynamic attributes of the graph nodes (e.g., load factor, temperature, and status code) are updated, forming a strong "device-data" mapping. Based on the electrical drawings of the airport's high-voltage, medium-voltage, and low-voltage power distribution systems, the connection relationships between transformers, switchgear, and cables are extracted (e.g., "transformer T1 → switchgear G1 → cable L1"). For the airport's heating / cooling system, extract the connection relationships between boilers, heat exchange stations, and pipelines (e.g., "Boiler B1 → Pipeline P1 → Terminal Air Conditioning Unit A1"). Integrate the electrical and thermal topologies into a unified multi-energy flow topology, supporting joint analysis across energy types (electricity, heat, and cooling). Construct a topological hierarchy based on voltage levels or energy types. Define physical parameters (e.g., maximum cable current capacity, pipeline flow limit) and operational constraints (e.g., cable temperature limit, pipeline pressure threshold) for each connection edge. Add directional arrows (e.g., current direction, heat flow direction) to energy transmission paths to form directed connections. Capture physical device state change events (e.g., switch opening and closing, device startup and shutdown, fault alarms) through sensor data and device logs (e.g., PLC control records). Add a precise timestamp (with millisecond accuracy) to each state change event and associate it with the corresponding graph node or connection edge. Map device state change events to the topology graph (e.g., switch opening invalidates the connection edge, device failure changes the graph node attributes). Store timestamped topology snapshots (node ​​and edge states) in a time series database, supporting queries by time range (e.g., querying topology changes within a certain time period). Based on time series data, generate dynamic topology maps with timelines (e.g., animated GIFs or interactive timelines), allowing users to drag the time slider to view topology states at different moments.

[0066] Physical devices such as transformers, energy storage cabinets, and photovoltaic panels are converted into graph nodes with spatial coordinates and bound to real-time sensor data streams, achieving a three-in-one modeling of device physical, data, and spatial attributes, eliminating the disconnect between devices and data in traditional energy management. Directed connections between devices are constructed to clarify energy transmission paths and directions, quantify hierarchical relationships and capacity constraints, and provide structured input for load forecasting and supply-demand balancing analysis. Device state changes (such as on / off status and load factor) are recorded and a timestamped topology map is generated to ensure data timeliness. Timestamps are used to trace the topology and device status at any moment, assisting in fault cause analysis and responsibility assignment. The topology map provides device association weights for the load forecasting model, reducing forecast errors and eliminating redundant energy storage and backup power configurations. Based on topology and device status data, fault sources can be quickly located, significantly shortening response time.

[0067] Optionally, predicting the energy load demand of the airport at different time periods according to the energy topology map includes:

[0068] Extracting node dynamic parameters and device association weights from the energy topology map, and building a multidimensional feature matrix based on historical load patterns;

[0069] Inputting the equipment status data, flight dynamics, meteorological parameters and the multidimensional feature matrix in the energy topology map into a long short-term memory network to generate a benchmark forecast curve for load demand by time period;

[0070] Based on the failure probability of graph nodes, meteorological volatility and the benchmark prediction curve, the Monte Carlo method is used to simulate the preset abnormal scenarios and output the energy load demand in different time periods.

[0071] Dynamic parameters of graph nodes (such as transformers, energy storage cabinets, and photovoltaic panels) are extracted in real time from the energy topology graph, including real-time power (active / reactive power, current, and voltage), state variables (switch state, SOC state of charge, and temperature), and operating efficiency (such as transformer load factor and photovoltaic conversion efficiency). By integrating sensor data streams (such as smart meters and current transformers) into the graph, dynamic mapping of device parameters to the topology is achieved (e.g., when a node's power is abnormal, its position in the topology and associated devices can be quickly located). Based on the directed edges of the topology graph, the energy transmission dependencies between devices are quantified (e.g., the proportion of power supplied by node A to node B). Historical data is used to mine the synergies of device operating modes (e.g., the correlation between terminal lighting and flight density, and the correlation between air conditioning systems and outdoor temperature) to construct a device association weight matrix. Time-based (e.g., 15-minute load curves) are extracted from historical load data and, combined with static attributes of the topology graph (e.g., device capacity and spatial location), historical load pattern labels (e.g., "weekday morning peak" and "holiday night low") are generated. Node dynamic parameters, equipment association weights, and historical load patterns are integrated into a multidimensional feature matrix, including equipment status features (such as real-time power and load factor), topological structural features (such as node degree centrality and betweenness centrality), and spatiotemporal features (such as time of day, season, and weekdays / holidays). Real-time equipment status (such as energy storage SOC and transformer temperature) is extracted from the topological map to quantify equipment health and operational efficiency. The system interfaces with the airport operations system or air traffic control bureau to obtain information related to flight schedules (takeoff and landing times, aircraft types, and passenger capacity), actual operating status (delays / cancellations), and energy consumption of ground support equipment (boarding bridges and jet bridge lighting). Satellite cloud images and ground-based meteorological stations are integrated to construct a spatiotemporal gridded meteorological database. Equipment status, flight dynamics, meteorological parameters, and the topological feature matrix are concatenated into an input vector, which serves as the time-series input sequence for the LSTM. A two-layer LSTM network is used to capture the temporal dependencies among equipment status, flight dynamics, and meteorological parameters (e.g., the impact of a three-hour flight delay on load). A self-attention module is introduced into the LSTM output layer to dynamically assign weights to different features (such as flight density and sunlight intensity) for load forecasting, improving the ability to capture key features. A fully connected layer outputs time-segmented load forecasts for the next 24-72 hours (at 15-minute intervals) to generate a baseline forecast curve (including expected load values ​​and standard deviations). Based on historical equipment failure data recorded in the topology map (such as transformer failure rates and photovoltaic panel degradation rates), a list of faulty devices and a distribution of failure times (e.g., using a Poisson distribution to simulate sudden failures) are generated. Incorporating historical volatility in meteorological parameters (such as wind speed standard deviation and sunlight intensity coefficient of variation), extreme weather scenarios are generated (e.g., heavy rain causing a sudden drop in photovoltaic output and high temperatures causing a surge in air conditioning load). Scenarios such as flight schedule changes and ground support equipment failures (e.g., boarding bridge power failure) are simulated to quantify their impact on load.Generate multiple preset abnormal scenarios, covering multiple combinations such as equipment failure, sudden weather changes, and flight disruptions. Each scenario is input into the LSTM model to generate time-based load forecasts. Combined with the baseline forecast curve, the load deviation (e.g., the change in peak load under the failure scenario) is calculated. The distribution of load forecast values ​​for each time period (e.g., mean, median, 95% confidence interval) is calculated, and a load probability density function is generated to quantify the load fluctuation range under different abnormal scenarios. Output is provided for the expected load values ​​and probability distribution for the next 24-72 hours (e.g., "Expected load value 1200kW from 08:00-09:00, 95% confidence interval [1000, 1400]kW"). Based on the probability of load fluctuation, high-risk time periods (e.g., periods with a load peak probability >80%) and areas (e.g., Terminal 2) are identified to assist operations personnel in developing response strategies.

[0072] The system extracts dynamic parameters of device nodes (such as real-time power and load factor) and device-associated weights from the energy topology map, quantifying the coupling relationships between physical devices. This transforms traditional single-time series forecasting into spatiotemporal correlation feature mining, reducing forecasting errors. By combining historical load patterns (such as seasonal fluctuations and holiday effects) with real-time data (such as flight dynamics and meteorological parameters), it constructs a three-dimensional feature matrix encompassing spatial topology, time series, and external influencing factors, enhancing the forecasting model's adaptability to complex scenarios. Leveraging the LSTM recurrent neural network architecture, it performs minute-level time series modeling on device status data (such as energy storage system SOC changes), flight dynamics (such as takeoff and landing times, aircraft passenger capacity), and meteorological parameters (such as temperature and light intensity) from the energy topology map. This model captures both short-term fluctuations in load (such as power consumption surges during peak flight periods) and long-term trends (such as seasonal load growth). It then outputs a time-based load demand benchmark curve (such as hourly load values) for the next 24-72 hours, with a temporal resolution of 15 minutes, providing a highly accurate basis for energy scheduling decisions. Based on node failure probabilities (such as historical transformer failure rates and cable aging coefficients) and meteorological volatility (such as wind speed standard deviation and cloud cover) in the energy topology map, the Monte Carlo method generates multiple pre-defined abnormal scenarios (such as sudden equipment failures, extreme weather, and flight delays), covering over 95% of potential risk combinations. Load demand changes are simulated for each scenario, generating load demand probability distributions (such as 95% confidence intervals) for different time periods. Output is a load demand probability cloud map that includes expected value, variance, and risk level, providing quantitative support for emergency plan development.

[0073] Optionally, the predicting of light intensity and wind speed according to the meteorological data, and determining the supply of renewable energy in different time periods according to the light intensity and wind speed includes:

[0074] Based on satellite cloud images and ground meteorological station data, the temporal fluctuation characteristics of light intensity and the spatial distribution of wind fields in the future preset time period are extracted to construct a spatiotemporal grid meteorological parameter matrix;

[0075] Predicting the probability distribution of light intensity based on the spatiotemporal gridded meteorological parameter matrix and generating a joint probability density function of renewable energy supply by combining wind turbine power curves and turbulence intensity coefficients;

[0076] The time-division output power of renewable energy is obtained according to the equipment parameters in the energy topology map and the joint probability density function.

[0077] The integration of satellite cloud imagery (covering a wide range of cloud motion patterns and reflectivity) with ground-based meteorological station data (measured wind speed, wind direction, and solar radiation intensity) addresses the challenges of insufficient spatial resolution (satellite) or poor temporal continuity (ground-based station) of a single data source. The airport coverage area is divided into 1km×1km grid cells. Within each grid, temporal fluctuations in solar intensity (e.g., a 15-minute radiation value curve) and the spatial distribution of wind fields (e.g., wind speed vector and wind direction angle) are extracted. A 4D space-time matrix (three-dimensional space + time) is constructed to support refined modeling of meteorological parameters. Based on historical meteorological data and cloud motion models, a Gaussian mixture model or deep generative adversarial network is used to predict the probability distribution of solar intensity (e.g., a 95% confidence interval of [300-800] W / m²) for each grid cell over a future period (e.g., 24 hours) and quantify uncertainty. Combining the wind turbine power curve (input wind speed-output power mapping) with the turbulence intensity coefficient (reflecting the impact of wind speed fluctuations on power generation efficiency, such as a 10% power reduction when turbulence intensity >15%), a wind power output probability density function is constructed to describe the discreteness of power generation under different wind speed conditions. Using Copula theory or Monte Carlo sampling, the marginal probability distributions of light intensity and wind power generation are coupled into a joint probability density function, reflecting the complementarity of the two energy sources (e.g., strong daytime light with weak wind and strong nighttime wind with weak light). Physical parameters of devices such as photovoltaic arrays (e.g., rated power, tilt angle, orientation) and wind turbines (e.g., cut-in / cut-out wind speed, rated speed) are extracted from the energy topology map and combined with real-time status (e.g., dust accumulation rate of photovoltaic panels and wear of wind turbine blades) to construct a device-level energy conversion model. The joint probability density function is combined with the equipment parameters, and Monte Carlo integration or dynamic programming algorithms are used to calculate the renewable energy supply for each time period (e.g., 1 hour) (e.g., the mean ± standard deviation of photovoltaic output power, and the range of wind turbine output power). The output result is a time-based power series with confidence intervals (e.g., photovoltaic: 8:00-9:00, 450±50kW; wind turbine: 22:00-23:00, 600-800kW).

[0078] Based on the probability distribution of sunlight intensity and wind turbine power curves, a joint probability density function (JPDF) of photovoltaic and wind power output is constructed to quantify energy supply fluctuations under different weather conditions (e.g., the complementary effect when sunlight is low but wind speed is high). This JPDF identifies high-risk scenarios (e.g., persistent rainy weather combined with low wind speeds) and outputs the probability of insufficient renewable energy supply (e.g., the probability of a supply shortfall exceeding 30% for three consecutive hours is less than 5%), providing a warning threshold for triggering backup power. By combining device parameters from the energy topology map (e.g., photovoltaic module conversion efficiency, wind turbine hub height, and rotor diameter), the JPDF is mapped to time-segmented output power (e.g., expected hourly photovoltaic / wind power output, maximum / minimum power output) with minute-level temporal resolution (e.g., predicted values ​​updated every 15 minutes). For airport distributed photovoltaic arrays and wind turbine clusters, the output power forecast dynamically adjusts the unit start / stop strategies (e.g., prioritizing photovoltaic panels in areas with ample sunlight) to improve overall power generation efficiency. Based on the joint probability distribution of output power, a reliability index for renewable energy supply is calculated to assist in decision-making regarding whether to activate energy storage systems or purchase power from the grid. By combining flight dynamics with energy demand forecasts, a time-segmented source-grid-load-storage coordinated scheduling strategy is generated (e.g., prioritizing photovoltaic consumption during the peak sunlight period at noon and energy storage charging during strong winds at night), thereby improving the utilization rate of renewable energy.

[0079] Optionally, determining the time-sharing energy supply and demand balance probability according to the energy load demand and the supply includes:

[0080] Inputting the energy load demand and the supply into a Bayesian network, constructing a supply-demand joint probability model, and generating a time-divided supply-demand gap probability density function;

[0081] Simulating the preset scenario based on the Monte Carlo method, and calculating the expected value and probability of the supply and demand gap in each time period according to the supply and demand gap probability density function of the time period;

[0082] The probabilities are mapped to the digital twin of the airport, and a visual heat map is generated based on operational constraints to identify time periods and areas where the intervention priority is greater than a threshold.

[0083] The energy load demand forecast (including probability distribution) and renewable energy supply forecast (including joint probability density function) are used as input variables to construct a Bayesian network topology. The causal relationship between load demand and supply is quantified (e.g., a decrease in sunlight intensity leads to a decrease in photovoltaic supply, while an increase in flight volume leads to a surge in load). The joint probability distribution P(L, S) is calculated using the Bayesian formula, where L is the load demand and S is the supply. Uncertainties in load demand (e.g., flight delays, sudden weather changes) and supply (e.g., cloud cover, wind speed fluctuations) are propagated through the probabilistic chain rule to generate a time-based supply-demand gap probability density function. This time-based supply-demand gap probability distribution (e.g., normal distribution or mixed Gaussian distribution) is generated for the next 24-72 hours. Key output parameters include expected value (average value of the supply-demand gap for each time period), variance (range of the gap fluctuation), and probability threshold (probability of the gap exceeding the threshold). Based on the supply-demand gap probability density function, Latin Hypercube Sampling (LHS) is used to generate 1,000-5,000 pre-set scenarios, covering over 95% of potential risk combinations. Based on operational constraints in the airport digital twin (e.g., energy storage system charge and discharge rate limits, backup power supply startup time, and grid power purchase price tiers), scenario weights are dynamically adjusted (e.g., when electricity purchase costs are high during peak hours, the simulation probability of insufficient energy storage is increased). For each scenario, the expected supply and demand gap and gap probability are calculated to generate a time-based risk matrix. Based on the risk matrix, time periods are categorized as high risk (red), medium risk (yellow), and low risk (green), and the risk level and gap probability for each time period are output. The supply and demand gap probabilities are mapped to the 3D model of the airport digital twin (e.g., terminals, runways, and aprons), generating heat maps by region and time period (e.g., red areas indicate high-risk time periods and areas). The heat map can be viewed by scrolling along the timeline, overlaying relevant information such as flight dynamics (e.g., takeoffs and landings) and equipment status (e.g., energy storage SOC). Intervention priority thresholds are set (e.g., gap probability >80% and expected value >3MW are considered high priority). For high-priority periods and areas, intervention suggestions are automatically generated (such as starting backup generators, adjusting flight charging plans, and optimizing energy storage charging and discharging strategies) and pushed to the energy management system and operator terminals.

[0084] Through probabilistic analysis and visual early warning, response time to extreme scenarios is significantly shortened, ensuring power continuity for critical facilities (such as navigation systems). Optimizing backup power and energy storage scheduling strategies reduces redundancy costs (for example, the traditional requirement of 20% reserve capacity has been reduced to 12% after optimization). This also reduces equipment losses caused by over-response (for example, by reducing the number of false starts of diesel generators). While ensuring safety, this maximizes renewable energy consumption, reduces grid electricity purchase costs, and helps airports achieve their near-zero carbon operation goals.

[0085] Optionally, determining a preliminary energy management plan according to the cost parameter, the time-sharing energy supply and demand balance probability, and equipment operation constraints includes:

[0086] Based on the energy price data, the unit energy cost of different energy types in each time period is calculated, and based on the unit energy cost, the charging and discharging cost of the energy storage system, and the start-up and shutdown cost of the backup generator, an energy cost parameter matrix for each time period is constructed;

[0087] Taking minimizing total energy cost as the objective function, a multi-objective optimization model is constructed according to the energy cost parameter matrix, the time-sharing energy supply and demand balance probability and equipment operation constraints;

[0088] The multi-objective optimization model is solved to obtain the energy storage system charge and discharge timing, the backup generator start and stop plan, and the power purchase ratio of the power grid to form a preliminary energy management plan.

[0089] Based on time-of-use electricity prices (e.g., peak-valley and peak-peak electricity prices), calculate the grid purchase price for each period (e.g., 0.8 yuan / kWh (peak), 0.3 yuan / kWh (valley)). Calculate the marginal cost of photovoltaic / wind power (including equipment depreciation and operation and maintenance costs, e.g., 0.15 yuan / kWh). Convert charging and discharging losses (e.g., 95% charge and discharge efficiency) and battery life degradation (e.g., cycle limit) into unit costs (e.g., 0.02 yuan / kWh charge and discharge costs). Quantify backup power costs: diesel generator start-up and shutdown costs (including fuel, startup wear and tear, and manual intervention costs, e.g., a single start-up cost of 500 yuan); gas turbine response costs (including fuel, emission taxes, and rapid start-up and shutdown losses, e.g., 0.5 yuan / kWh). Generate a 24×N matrix (N is the energy type, e.g., grid, photovoltaic, or energy storage), with each row corresponding to a time period and each column corresponding to the energy cost. Dynamic updates are supported (e.g., real-time electricity price fluctuations). Economic objective: minimize total cost. Reliability objective: minimize the expected supply-demand gap. Constraints include equipment operation constraints and supply-demand balance constraints. Equipment operation constraints include: energy storage system SOC (state of charge) range constraints (e.g., 20%-90%), charge and discharge power limits (e.g., ±500kW), daily cycle limits (e.g., ≤2 times); minimum standby generator operating time (e.g., 1 hour), start interval (e.g., ≥4 hours), and maximum continuous operating time (e.g., ≤8 hours); and maximum power limits for grid-purchased electricity (e.g., transformer capacity limits). Supply-demand balance constraints include: energy supply must meet load demand probability thresholds in each time period (e.g., a gap of ≤5% under a 95% confidence level). Mixed integer linear programming is used to jointly optimize discrete variables such as energy storage start-up and shutdown, standby power supply switches, and continuous variables such as charge and discharge power and power purchase ratio. Use a solver (such as Gurobi or CPLEX) or an open-source framework (such as Pyomo) for global optimization to determine hourly charge and discharge power (e.g., charging energy storage from 2:00 PM to 4:00 PM, discharging from 8:00 PM to 10:00 PM), diesel generator startup times (e.g., shutting down during the early morning peak load period and starting up 30 minutes before the peak load period), and the proportion of grid-purchased electricity to total load during each time period (e.g., 80% during off-peak hours). Generate a Gantt chart (with time on the horizontal axis and device status on the vertical axis, such as energy storage charging / discharging and backup power supply operating / standby) and a cost-risk curve (with total cost on the horizontal axis and the probability of a supply-demand gap on the vertical axis, marking the Pareto frontier).

[0090] The charging and discharging costs of the energy storage system (such as battery life loss and efficiency loss) and the startup and shutdown costs of the backup generator (such as diesel consumption, engine wear, and carbon emissions) are quantified to avoid decision-making biases caused by traditional methods that ignore hidden costs (e.g., overcharging and discharging due to failure to consider energy storage life costs). Minimizing total energy cost is the core objective function, integrating supply and demand balance probability constraints, equipment operation constraints, and green energy priority constraints. Using the analytic hierarchy process or entropy weighting method, target weights are dynamically assigned based on airport operational priorities (e.g., a 60% weight for economy, a 30% weight for reliability, and a 10% weight for green energy) to achieve a Pareto optimal solution. A hybrid branch-and-bound genetic algorithm solver is employed to address nonlinear constraints (e.g., the nonlinear relationship between energy storage charging and discharging efficiency and SOC) and discrete variables (e.g., the startup and shutdown status of the backup generator) in the model. This ensures global optimality while keeping the solution time to minutes (e.g., solving a 24-hour scheduling plan within 10 minutes). Through a rolling optimization mechanism, the latest data such as supply and demand balance probability, energy prices, and equipment status are fed back to the model, enabling minute-level dynamic adjustment of the scheduling plan and improving adaptability to emergency scenarios (such as sudden load changes caused by flight delays).

[0091] Optionally, adjusting the preliminary energy management plan based on the difference between the first energy performance data and the second energy performance data includes:

[0092] Comparing and analyzing the first energy execution data with the second energy execution data, calculating differences between the first energy execution data and the second energy execution data in terms of operating parameters of each physical device, energy usage parameters of each region, and energy cost parameters, and determining a degree of difference based on the differences;

[0093] determining a target factor causing the difference according to the degree of the difference and the amount of change in the multi-source data of the airport;

[0094] The preliminary energy management plan is adjusted according to the target factors.

[0095] The execution data is broken down into three core dimensions for comparison:

[0096] Physical equipment operating parameters: energy storage system SOC (state of charge) deviation, backup generator output power fluctuation, photovoltaic inverter conversion efficiency changes, etc.

[0097] Regional energy usage parameters: differences in power loads across the terminal (check-in area, departure lounge, baggage claim area), and energy consumption deviations of apron equipment (jet bridges, boarding stairs) (e.g., power consumption in Area A exceeds plan by 12%).

[0098] Energy cost parameters: deviations from the planned actual grid electricity purchase cost, diesel consumption cost, and energy storage charging and discharging loss cost (e.g., the actual electricity purchase price during peak hours is 15% higher than planned).

[0099] The difference value is calculated by combining the standardized Z-Score with the weighted Euclidean distance. The difference threshold is dynamically adjusted according to the airport operation stage (such as peak / off-peak flight periods) and energy market volatility (such as when the peak-to-valley difference in electricity prices is >50%). For example, during highly sensitive periods (such as periods with dense flights): the equipment parameter deviation threshold is ±3%, and the cost deviation threshold is ±5%; during less sensitive periods (such as night flight intervals): the equipment parameter deviation threshold is ±8%, and the cost deviation threshold is ±10%. The difference value is divided into three levels of warning: green level (difference < threshold): no plan adjustment is required; yellow level (threshold ≤ difference < 2 times the threshold): trigger local optimization of the plan; red level (difference ≥ 2 times the threshold): initiate a comprehensive reconstruction of the plan. Based on the dual engines of machine learning causal inference (such as Pearl causal diagrams and counterfactual reasoning) and physical mechanism analysis, the target factors that cause the difference are identified:

[0100] Meteorological data deviation: For example, actual sunlight intensity is 20% lower than the predicted value (cloud cover exceeds expectations), resulting in reduced photovoltaic power generation;

[0101] Abnormal equipment status: For example, the internal resistance of the energy storage system battery pack increases by 15%, resulting in a decrease in charging and discharging efficiency;

[0102] Dynamic changes in flight conditions: For example, the temporary addition of three wide-body aircraft (APUs consume a lot of energy), which results in a surge in apron power load;

[0103] Sudden changes in market prices: For example, if the real-time electricity price increases by 30% compared to the planned value, it will lead to an overspending on electricity purchase costs.

[0104] Calculate the contribution of each factor to the difference value, for example:

[0105] Contribution of meteorological deviation: 40% (insufficient sunlight causes 1500kWh reduction in photovoltaic power generation);

[0106] Contribution of equipment failure: 30% (reduced energy storage efficiency leads to an increase in costs of 800 yuan);

[0107] Contribution of flight changes: 25% (the increase in wide-body aircraft resulted in an excess of 1,200 kWh in electricity consumption);

[0108] Contribution of market fluctuations: 5% (a small fluctuation in electricity prices affects costs by 200 yuan).

[0109] Sort by contribution in descending order, giving priority to high-contribution factors. Match optimization strategies based on target factor types:

[0110] Weather deviations: Activate backup power sources (such as diesel generators) or adjust energy storage charging and discharging plans (such as delaying discharge to cope with subsequent shortfalls);

[0111] Equipment failure: Isolate the faulty equipment (e.g., deactivate the abnormal battery pack) and reallocate the output of other equipment (e.g., call on adjacent energy storage units);

[0112] Flight dynamics: Dynamically adjust the operational priority of apron equipment (e.g. prioritize jet bridge power supply and shut down non-critical lighting);

[0113] Market fluctuations: Switching electricity purchasing channels (such as switching from the real-time electricity price market to the contract market) or calling on distributed power sources (such as increasing self-generation and self-use of photovoltaic power).

[0114] Using the model predictive control framework, the following steps are performed in a 15-minute cycle:

[0115] Perform data collection: obtain the latest equipment status, load, and cost data;

[0116] Variance calculation and root cause analysis: Identify new variances based on the comparison of current data with the latest plan;

[0117] Plan revision: Adjust control parameters (e.g., energy storage SOC target value ±2%) based on the root cause and generate a new plan;

[0118] Instruction issuance: Push the corrected charge and discharge power, power purchase ratio and other instructions to the device controller.

[0119] Simulation verification and safety check: Before the plan is issued, a 10-second rapid simulation is performed through the digital twin to verify key indicators such as the adjusted supply and demand balance probability (such as gap probability <1%) and equipment overload risk (such as transformer load rate <85%).

[0120] By combining changes in airport data from multiple sources (such as weather conditions, flight dynamics, equipment failure records, and electricity price fluctuations), a causal graph model or random forest feature importance analysis is used to quantify the contribution of each factor to execution variance. For complex scenarios (such as typhoons causing simultaneous fluctuations in photovoltaic and wind power), principal component analysis (PCA) and SHAP value interpretation are used to separate the independent and interactive effects of various factors, avoiding the biased attribution of a single factor. For equipment or areas where variance exceeds a high-risk threshold, emergency response plans (such as activating backup diesel generators for grid connection and initiating emergency load rationing) are automatically triggered, reducing the impact of deviations from hours required by traditional manual response to minutes. Based on execution variance data, equipment failure probability models (such as correcting for energy storage battery aging rates) and load forecast error distributions (such as adjusting the terminal air conditioning load fluctuation coefficient) are updated to dynamically improve the robustness of subsequent planning. Real-time correction reduces the additional costs associated with deviations, further reducing total energy costs compared to traditional methods. Improve the accuracy of renewable energy consumption, optimize the charging and discharging efficiency of energy storage systems, reduce redundant operation of backup power supplies, and achieve overall energy efficiency improvement of airport energy systems.

[0121] This embodiment also discloses a system for airport energy balance management. Figure 2 This is a module diagram of a system for airport energy balance management disclosed in an embodiment of the present application, such as Figure 2 As shown, the system includes a graph module 201, a prediction module 202, a planning module 203, and an adjustment module 204, wherein:

[0122] A mapping module 201 is configured to acquire multi-source data and construct an energy topology map based on the multi-source data, wherein the multi-source data includes flight data, weather data, energy usage data of various areas within the airport, and energy price data;

[0123] Prediction module 202 is configured to predict the energy load demand of the airport at different time periods based on the energy topology map, predict the light intensity and wind speed based on the meteorological data, determine the supply of renewable energy at different time periods based on the light intensity and wind speed, and determine the time-sharing energy supply and demand balance probability based on the energy load demand and the supply;

[0124] a planning module 203 configured to determine cost parameters based on the energy price data, and determine a preliminary energy management plan based on the cost parameters, the time-of-use energy supply and demand balance probability, and equipment operation constraints, wherein the preliminary energy management plan includes a charge and discharge sequence of the energy storage system, a start and stop plan for the backup generator, and a proportion of power purchased from the power grid;

[0125] The adjustment module 204 is configured to execute the preliminary energy management plan through the digital twin of the airport to obtain first energy execution data, obtain second energy execution data at the current moment, and adjust the preliminary energy management plan according to the difference between the first energy execution data and the second energy execution data.

[0126] Optionally, the atlas module 201 is configured to:

[0127] Based on the airport's building information model and the coordinates of the physical equipment, the physical equipment is converted into graph nodes with spatial attributes, and the real-time sensor data stream is bound to establish a mapping relationship between the physical equipment and the data;

[0128] According to the energy transmission path, the directed connection edges between the physical devices are constructed, and the hierarchical relationship and capacity constraints are defined to form an initial static topology skeleton;

[0129] The state changes of the physical devices are recorded in a time series database, and an energy topology map with a time stamp is generated according to the state changes and the initial static topology skeleton.

[0130] Optionally, the prediction module 202 is configured to:

[0131] Extracting node dynamic parameters and device association weights from the energy topology map, and building a multidimensional feature matrix based on historical load patterns;

[0132] Inputting the equipment status data, flight dynamics, meteorological parameters and the multidimensional feature matrix in the energy topology map into a long short-term memory network to generate a benchmark forecast curve for load demand by time period;

[0133] Based on the failure probability of graph nodes, meteorological volatility and the benchmark prediction curve, the Monte Carlo method is used to simulate the preset abnormal scenarios and output the energy load demand in different time periods.

[0134] Optionally, the prediction module 202 is configured to:

[0135] Based on satellite cloud images and ground meteorological station data, the temporal fluctuation characteristics of light intensity and the spatial distribution of wind fields in the future preset time period are extracted to construct a spatiotemporal grid meteorological parameter matrix;

[0136] Predicting the probability distribution of light intensity based on the spatiotemporal gridded meteorological parameter matrix and generating a joint probability density function of renewable energy supply by combining wind turbine power curves and turbulence intensity coefficients;

[0137] The time-division output power of renewable energy is obtained according to the equipment parameters in the energy topology map and the joint probability density function.

[0138] Optionally, the prediction module 202 is configured to:

[0139] Inputting the energy load demand and the supply into a Bayesian network, constructing a supply-demand joint probability model, and generating a time-divided supply-demand gap probability density function;

[0140] Simulating the preset scenario based on the Monte Carlo method, and calculating the expected value and probability of the supply and demand gap in each time period according to the supply and demand gap probability density function of the time period;

[0141] The probabilities are mapped to the digital twin of the airport, and a visual heat map is generated based on operational constraints to identify time periods and areas where the intervention priority is greater than a threshold.

[0142] Optionally, the planning module 203 is configured to:

[0143] Based on the energy price data, the unit energy cost of different energy types in each time period is calculated, and based on the unit energy cost, the charging and discharging cost of the energy storage system, and the start-up and shutdown cost of the backup generator, an energy cost parameter matrix for each time period is constructed;

[0144] Taking minimizing total energy cost as the objective function, a multi-objective optimization model is constructed according to the energy cost parameter matrix, the time-sharing energy supply and demand balance probability and equipment operation constraints;

[0145] The multi-objective optimization model is solved to obtain the energy storage system charge and discharge timing, the backup generator start and stop plan, and the power purchase ratio of the power grid to form a preliminary energy management plan.

[0146] Optionally, the adjustment module 204 is configured to:

[0147] Comparing and analyzing the first energy execution data with the second energy execution data, calculating differences between the first energy execution data and the second energy execution data in terms of operating parameters of each physical device, energy usage parameters of each region, and energy cost parameters, and determining a degree of difference based on the differences;

[0148] determining a target factor causing the difference according to the degree of the difference and the amount of change in the multi-source data of the airport;

[0149] The preliminary energy management plan is adjusted according to the target factors.

[0150] It should be noted that the above embodiments provide devices that implement their functions using only the division of the above functional modules as examples. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0151] This embodiment also discloses an electronic device, referring to Figure 3 The electronic device may include: at least one processor 301 , at least one communication bus 302 , a user interface 303 , a network interface 304 , and at least one memory 305 .

[0152] The communication bus 302 is used to implement the connection and communication between these components.

[0153] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0154] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0155] The processor 301 may include one or more processing cores. Using various interfaces and circuits, the processor 301 connects to various components within the server. It executes instructions, programs, code sets, or instruction sets stored in the memory 305, as well as accesses data stored in the memory 305, to perform various server functions and process data. Optionally, the processor 301 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 301 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing content displayed on the display screen; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 301 but implemented as a separate chip.

[0156] Among them, the memory 305 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also be optionally at least one storage device located away from the aforementioned processor 301. As Figure 3 As shown, the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of a method for airport energy balance management.

[0157] exist Figure 3 In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 301 can be used to call an application program storing a method for airport energy balance management in the memory 305. When executed by one or more processors 301, the electronic device executes one or more methods in the above-mentioned embodiments.

[0158] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for this application.

[0159] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0160] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely schematic, such as the division of units, which is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interface, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0161] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0162] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0163] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of this application, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory 305 and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of this application. The aforementioned memory 305 includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a magnetic disk, or an optical disk.

[0164] The above is only an exemplary embodiment of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any variations, uses or adaptive changes of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary technical means in the technical field that are not recorded in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A method for airport energy balance management, characterized in that: Applied to an energy management platform, the method includes: Acquiring multi-source data, including flight data, weather data, energy usage data for various areas within the airport, and energy price data, and constructing an energy topology map based on the multi-source data; Predicting the energy load demand of the airport at different time periods based on the energy topology map, predicting the light intensity and wind speed based on the meteorological data, determining the supply of renewable energy at different time periods based on the light intensity and wind speed, and determining the time-sharing energy supply and demand balance probability based on the energy load demand and the supply; Determining cost parameters based on the energy price data, and determining a preliminary energy management plan based on the cost parameters, the time-sharing energy supply and demand balance probability, and equipment operation constraints, the preliminary energy management plan including the energy storage system charge and discharge sequence, the backup generator start and stop plan, and the power purchase ratio of the power grid; Executing the preliminary energy management plan through the digital twin of the airport to obtain first energy execution data, obtaining second energy execution data at the current moment, and adjusting the preliminary energy management plan based on the difference between the first energy execution data and the second energy execution data. The constructing of an energy topology map based on the multi-source data includes: Based on the airport's building information model and the coordinates of the physical equipment, the physical equipment is converted into graph nodes with spatial attributes, and the real-time sensor data stream is bound to establish a mapping relationship between the physical equipment and the data; According to the energy transmission path, the directed connection edges between the physical devices are constructed, and the hierarchical relationship and capacity constraints are defined to form an initial static topology skeleton; Record the state changes of the physical devices through a time series database, and generate a time-stamped energy topology map based on the state changes and the initial static topology skeleton. The predicting of the energy load demand of the airport at different time periods according to the energy topology map includes: Extracting node dynamic parameters and device association weights from the energy topology map, and building a multidimensional feature matrix based on historical load patterns; Inputting the equipment status data, flight dynamics, meteorological parameters and the multidimensional feature matrix in the energy topology map into a long short-term memory network to generate a benchmark forecast curve for load demand by time period; Based on the failure probability of graph nodes, meteorological volatility and the benchmark prediction curve, the Monte Carlo method is used to simulate the preset abnormal scenarios and output the energy load demand in different time periods.

2. The method for airport energy balance management according to claim 1, characterized in that: The step of predicting the light intensity and wind speed according to the meteorological data and determining the supply of renewable energy in different time periods according to the light intensity and wind speed includes: Based on satellite cloud images and ground meteorological station data, the temporal fluctuation characteristics of light intensity and the spatial distribution of wind fields in the future preset time period are extracted to construct a spatiotemporal grid meteorological parameter matrix; Predicting the probability distribution of light intensity based on the spatiotemporal gridded meteorological parameter matrix and generating a joint probability density function of renewable energy supply by combining wind turbine power curves and turbulence intensity coefficients; The time-division output power of renewable energy is obtained according to the equipment parameters in the energy topology map and the joint probability density function.

3. The method for airport energy balance management according to claim 1, characterized in that: Determining the time-sharing energy supply and demand balance probability according to the energy load demand and the supply includes: Inputting the energy load demand and the supply into a Bayesian network, constructing a supply-demand joint probability model, and generating a time-divided supply-demand gap probability density function; Simulating the preset scenario based on the Monte Carlo method, and calculating the expected value and probability of the supply and demand gap in each time period according to the supply and demand gap probability density function of the time period; The probabilities are mapped to the digital twin of the airport, and a visual heat map is generated based on operational constraints to identify time periods and areas where the intervention priority is greater than a threshold.

4. The method for airport energy balance management according to claim 1, characterized in that: Determining a preliminary energy management plan based on the cost parameters, the time-sharing energy supply and demand balance probability, and equipment operation constraints includes: Based on the energy price data, the unit energy cost of different energy types in each time period is calculated, and based on the unit energy cost, the charging and discharging cost of the energy storage system, and the start-up and shutdown cost of the backup generator, an energy cost parameter matrix for each time period is constructed; Taking minimizing total energy cost as the objective function, a multi-objective optimization model is constructed according to the energy cost parameter matrix, the time-sharing energy supply and demand balance probability and equipment operation constraints; The multi-objective optimization model is solved to obtain the energy storage system charge and discharge timing, the backup generator start and stop plan, and the power purchase ratio of the power grid to form a preliminary energy management plan.

5. The method for airport energy balance management according to claim 1, characterized in that: The adjusting the preliminary energy management plan according to the difference between the first energy performance data and the second energy performance data includes: Comparing and analyzing the first energy execution data with the second energy execution data, calculating differences between the first energy execution data and the second energy execution data in terms of operating parameters of each physical device, energy usage parameters of each region, and energy cost parameters, and determining a degree of difference based on the differences; determining a target factor causing the difference according to the degree of the difference and the amount of change in the multi-source data of the airport; The preliminary energy management plan is adjusted according to the target factors.

6. A system for airport energy balance management, characterized in that: It includes graph module, forecast module, planning module and adjustment module, among which: a mapping module configured to acquire multi-source data and construct an energy topology map based on the multi-source data, wherein the multi-source data includes flight data, weather data, energy usage data of various areas within the airport, and energy price data; a prediction module configured to predict the energy load demand of the airport at different time periods based on the energy topology map, predict the light intensity and wind speed based on the meteorological data, determine the supply of renewable energy at different time periods based on the light intensity and wind speed, and determine the time-sharing energy supply and demand balance probability based on the energy load demand and the supply; a planning module configured to determine cost parameters based on the energy price data, and determine a preliminary energy management plan based on the cost parameters, the time-sharing energy supply and demand balance probability, and equipment operation constraints, the preliminary energy management plan including a charge and discharge sequence of the energy storage system, a start and stop plan for the backup generator, and a proportion of power purchased from the power grid; an adjustment module configured to execute the preliminary energy management plan through the digital twin of the airport to obtain first energy execution data, obtain second energy execution data at a current moment, and adjust the preliminary energy management plan based on the difference between the first energy execution data and the second energy execution data; The atlas module is configured to: Based on the airport's building information model and the coordinates of the physical equipment, the physical equipment is converted into graph nodes with spatial attributes, and the real-time sensor data stream is bound to establish a mapping relationship between the physical equipment and the data; According to the energy transmission path, the directed connection edges between the physical devices are constructed, and the hierarchical relationship and capacity constraints are defined to form an initial static topology skeleton; Record the state changes of the physical devices through a time series database, and generate a time-stamped energy topology map based on the state changes and the initial static topology skeleton. The prediction module is configured to: Extracting node dynamic parameters and device association weights from the energy topology map, and building a multidimensional feature matrix based on historical load patterns; Inputting the equipment status data, flight dynamics, meteorological parameters and the multidimensional feature matrix in the energy topology map into a long short-term memory network to generate a benchmark forecast curve for load demand by time period; Based on the failure probability of graph nodes, meteorological volatility and the benchmark prediction curve, the Monte Carlo method is used to simulate the preset abnormal scenarios and output the energy load demand in different time periods.

7. An electronic device, characterized in that: It includes a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 5 is executed.

Images