Comprehensive energy optimization scheduling method for smart park

Abstract

The invention discloses a smart park comprehensive energy optimization scheduling method, and relates to the technical field of smart park energy management, and the method comprises the following steps: carrying out the real-time collection of multi-source data in advance; operating parameters of photovoltaic and wind power generation equipment, a gas turbine and an energy storage system and user-side power, heat and cold load data are obtained in real time through a distributed sensor network; constructing a load prediction model based on an LSTM neural network, inputting historical load data and meteorological data, and outputting a hourly load prediction value in the next 24 hours; and constructing a multi-target operation evaluation index. Through the constructed multi-target operation evaluation index system and the optimized scheduling model, the limitation of a traditional single target is broken through, collaborative optimization of energy consumption, environmental protection and reliability is realized, and the comprehensiveness and scientificity of energy management of the smart park are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of smart park energy management, and in particular to a method for optimizing and scheduling comprehensive energy in a smart park. Background Art

[0002] With the continued rise in global energy demand and the advancement of the "dual carbon" goals, smart parks, as key vehicles for energy consumption and management, face a pressing challenge: ensuring the scientific and efficient scheduling of energy. Traditional smart park energy scheduling methods focus on a single energy type, lacking the coordinated and optimized management of multiple energy sources such as electricity, heat, and cooling. This results in inefficient energy utilization and significant energy waste during transmission and use.

[0003] Currently, existing technologies often prioritize cost minimization or environmental protection as single optimization objectives, making it difficult to balance economic, environmental, and energy supply stability, and thus unable to meet the diverse development needs of smart parks. Furthermore, smart park energy systems are complex and volatile, and existing scheduling methods are insufficiently adaptable to the park's dynamically changing energy demands and equipment operating conditions. This makes it difficult to ensure the reliability of power, heating, and cooling, limiting the advancement of smart park energy management. Therefore, a method that comprehensively considers multiple factors and achieves efficient and optimized energy scheduling for smart parks is urgently needed.

[0004] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention

[0005] In response to the problems in related technologies, the present invention proposes a comprehensive energy optimization and scheduling method for smart parks, which is used to provide a comprehensive energy optimization and scheduling method for smart parks. By constructing a multi-objective optimization scheduling model and adopting mixed linear integer programming to solve it, the comprehensive energy system of the smart park is optimized in multiple dimensions such as energy consumption, environmental protection and reliability, thereby improving energy utilization efficiency, reducing operating costs, reducing environmental pollution, ensuring a stable and reliable energy supply, and providing technical support for the sustainable development of smart parks.

[0006] The technical solution of the present invention is achieved as follows:

[0007] A method for optimizing and scheduling comprehensive energy in a smart park, comprising the following steps:

[0008] Conduct real-time data collection from multiple sources in advance, and obtain operating parameters of photovoltaic, wind power generation equipment, gas turbines, and energy storage systems, as well as user-side electricity, heat, and cooling load data in real time through a distributed sensor network;

[0009] Build a load forecasting model based on the LSTM neural network, input historical load data and meteorological data, and output hourly load forecast values ​​for the next 24 hours;

[0010] Construct multi-objective operation evaluation indicators, including at least energy consumption evaluation indicators, environmental protection evaluation indicators, power supply reliability evaluation indicators and cooling supply reliability evaluation indicators;

[0011] An optimization scheduling model for the day-ahead economic scheduling stage is constructed with the core goal of minimizing operating costs. The objective function is constructed by integrating fuel costs and grid interaction costs. Power balance, equipment operation, and environmental protection are used as constraints to obtain the optimal scheduling solution through a hybrid linear integer programming solver.

[0012] Furthermore, the energy consumption evaluation index includes: total energy consumption of the park E total , expressed as:

[0013]

[0014] Among them, T is the number of time steps in the scheduling period, E gen,i (t) is the power generation of the i-th power generation equipment at time t, E grid,j (t) is the amount of electricity purchased by the jth grid interaction point at time t, E storage,k (t) is the discharge capacity of the kth energy storage device at time t;

[0015] Energy consumption per unit area E area , expressed as:

[0016]

[0017] Where S is the total area of ​​the park.

[0018] Furthermore, the environmental evaluation indicators include: carbon emissions C total , expressed as:

[0019]

[0020] Above, C i (t) is the carbon emission of the i-th power generation equipment at time t, C grid,j (t) is the carbon emissions of electricity purchased by the jth grid interaction point at time t;

[0021] Among them, the carbon emission calculation of the i-th power generation equipment is expressed as:

[0022] C i (t) = λ i ×F i (t);

[0023] Above, λ i is the carbon emission coefficient of the i-th power generation equipment, F i (t) is the fuel consumption of the i-th power generation equipment at time t.

[0024] Furthermore, the power supply reliability evaluation index includes: system power outage time T outage Calculation, expressed as:

[0025]

[0026] Above, T out (t) is the power outage time at time t, and T out (t) = 0, otherwise it is the duration of the power outage;

[0027] Among them, the number of power outages N outage , expressed as:

[0028]

[0029] Furthermore, the cooling reliability evaluation index includes: cooling temperature stability ΔT cool , expressed as:

[0030]

[0031] Above, T cool (t) is the actual cooling temperature at time t, T set_cool To set the cooling temperature;

[0032] Among them, the number of cooling interruptions N cool_outage , expressed as:

[0033]

[0034] Above, Q cool (t) is the cooling power at time t.

[0035] Furthermore, the objective function of the optimization scheduling model for the day-ahead economic scheduling stage is expressed as:

[0036]

[0037] Among them, fuel cost FC i (t), expressed as:

[0038]

[0039] Above, P i (t) is the power generated by the i-th power generation equipment at time t, F i (P i (t)) is the fuel consumption function of the power generation equipment, η i (P i (t)) is the efficiency function of power generation equipment, C fuel,i for fuel prices;

[0040] Among them, the grid interaction fee GIC j (t), expressed as:

[0041] GIC j (t) = P grid_buy,j (t)×C buy (t)-P grid_sell,j (t)×C sell (t);

[0042] Above, P grid_buy,j (t) is the power purchased by the jth grid interaction point at time t, P grid_sell,j (t) is the electricity sold, C buy (t) and C sell (t) The electricity purchase and sales prices respectively;

[0043] Among them, equipment maintenance cost MC k (t), expressed as:

[0044] MC k (t) = P k (t)×τ k ×C maintain,k ;

[0045] Above, P k (t) is the power of the kth device at time t, τk is the device operation time, C maintain,k is the maintenance cost coefficient;

[0046] Among them, the unit start-up and shutdown costs SSC p (t), expressed as:

[0047] SSC p (t)=Δu p (t)×C start,p ;

[0048] Above, Δu p (t)=|u p (t)-u p (t-1)|,u p (t) is the unit start-stop state variable, C start,p for startup costs;

[0049] Among them, heat sales revenue HSR r (t), expressed as:

[0050] CSR s (t) = Q cool_sell,s (t)×C cool_sell (t);

[0051] Above, Q cool_sell,s (t) is the heat purchased by the sth heat user at time t, Ccool_sell (t) is the heat price;

[0052] Among them, sales of cold revenue CSR s (t), expressed as:

[0053] CSR s (t) = Q cool_sell,s (t)×C cool_sell (t);

[0054] Above, Q cool_sell,s (t) is the cooling capacity purchased by the sth cooling user at time t, C cool_sell (t) is the cold price.

[0055] Furthermore, the construction of the optimization scheduling model for the day-ahead economic scheduling stage further includes: calibrating constraint conditions, including power balance constraints, equipment operation constraints, and environmental protection constraints;

[0056] Among them, the power balance constraints include:

[0057] Power balance:

[0058]

[0059] Thermal balance:

[0060]

[0061] Cold force balance:

[0062]

[0063] Equipment operation constraints include:

[0064] Power generation equipment output constraints:

[0065]

[0066] Energy storage equipment constraints:

[0067]

[0068] SOC min ≤SOC(t)≤SOC max ;

[0069] 0≤P storage,in (t)≤P in_max ;

[0070] 0≤P storage,out (t)≤P out_max ;

[0071] Unit start and stop constraints:

[0072]

[0073] Environmental protection constraints include:

[0074] Carbon emission constraints:

[0075]

[0076] Pollutant emission constraints:

[0077]

[0078] Furthermore, the hybrid linear integer programming solver obtains the optimal scheduling plan, including: using CPLEX or Gurobi hybrid linear integer programming solver, solving through a branch and bound algorithm, outputting the optimal decision variables for the charging and discharging status of the energy storage equipment, the flexible load adjustment amount, and the start and stop and output status of the unit, to form a comprehensive energy optimization scheduling plan for the next 24 hours.

[0079] Beneficial effects of the present invention:

[0080] 1. The present invention realizes multi-objective collaborative optimization. By constructing a multi-objective operation evaluation index system and an optimization scheduling model, it breaks through the limitations of traditional single objectives, realizes the collaborative optimization of energy consumption, environmental protection and reliability, and improves the comprehensiveness and scientificity of smart park energy management.

[0081] 2. The present invention improves energy utilization efficiency by coordinating the scheduling of multiple energy devices, fully tapping the potential of renewable energy, rationally allocating energy storage equipment and flexible loads, reducing energy loss, significantly improving the energy utilization efficiency of the park, and effectively reducing total energy consumption and energy consumption per unit area. At the same time, environmental constraints prompt the park to give priority to the use of clean energy and high-efficiency environmentally friendly equipment, significantly reducing carbon emissions and pollutant emissions, helping the park achieve green and low-carbon development goals, enhancing the park's environmental competitiveness, and further achieving reliable energy supply guarantees. Based on the reliability evaluation indicators and constraints of power supply, heating, and cooling, the optimized scheduling plan can effectively reduce the number and duration of energy supply interruptions, ensure the stable operation of production and life of park users, and improve the quality of park energy services.

[0082] 3. The present invention achieves strong adaptability and versatility. By relying on real-time data collection and dynamic model construction, it can quickly respond to changes in park energy demand and equipment status. It is suitable for smart parks of different scales and energy structures and has broad promotion and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0084] Figure 1 It is a flow chart of a comprehensive energy optimization and scheduling method for a smart park according to an embodiment of the present invention. DETAILED DESCRIPTION

[0085] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention are within the scope of protection of the present invention.

[0086] According to an embodiment of the present invention, a method for comprehensive energy optimization and scheduling of a smart park is provided.

[0087] like Figure 1 As shown, the smart park comprehensive energy optimization scheduling method according to an embodiment of the present invention includes the following steps:

[0088] Step S1: Pre-collect multi-source data in real time, and obtain in real time through a distributed sensor network the power generated by the solar photovoltaic panel (accuracy ±5W), light intensity (resolution 0.1klx); wind turbine output power (accuracy ±10kW), wind speed (resolution 0.1m / s); gas turbine fuel consumption (accuracy ±0.1L / h), power generation efficiency (resolution 0.1%), lithium battery energy storage system state of charge (SOC, accuracy ±1%), charge and discharge power (accuracy ±5kW); flywheel energy storage system speed (accuracy ±10rpm) and charge and discharge efficiency (resolution 0.1%).

[0089] Specifically, in application, for photovoltaic / wind power generation monitoring, the following are included: installing high-precision power sensors (accuracy ±5W) and light intensity sensors (resolution 0.1klx) on the park's solar photovoltaic panel arrays to collect power generation and environmental parameters at a minute-level frequency; deploying power transmitters (accuracy ±10kW) and anemometers (resolution 0.1m / s) at wind turbines to monitor power generation status in real time. For gas turbine monitoring, electromagnetic flow meters (accuracy ±0.1L / h) are installed on the gas turbine fuel pipeline to simultaneously collect fuel consumption and power generation efficiency (obtained through the device's own sensors, resolution 0.1%). At the same time, for energy storage system monitoring, the battery management system (BMS) reads the state of charge (SOC, accuracy ±1%) and charge and discharge power (accuracy ±5kW) in real time. For flywheel energy storage: installing a speed sensor (accuracy ±10rpm) and an efficiency monitoring module (resolution 0.1%) to track charge and discharge dynamics. In addition, IoT terminals such as smart electricity meters, thermal imagers, and cooling meters are deployed to collect real-time data on user-side electricity / heat / cooling loads.

[0090] Step S2, constructing a load forecasting model, includes the following steps:

[0091] The 3σ rule was used to eliminate outliers and the historical load data were processed by Min-Max normalization (range [0,1]), including office, industrial, residential, and commercial loads;

[0092] A prediction model is constructed based on the LSTM neural network. The input layer is the load data and meteorological data (temperature / humidity / light) of the previous 48 hours, and the output layer is the hourly load forecast value for the next 24 hours.

[0093] The Adam optimizer is used for model training, the loss function is the root mean square error (RMSE), the validation set accounts for 20%, and the prediction accuracy index is:

[0094] Power load: RMSE < 3%, mean absolute percentage error (MAPE) < 5%;

[0095] Thermal load: RMSE < 5%, MAPE < 8%;

[0096] Cooling load: RMSE < 4%, MAPE < 6%.

[0097] In this technical solution, for the LSTM neural network prediction model, the input layer contains 48 hours of historical load data (electricity, heat, and cooling) and concurrent meteorological data (temperature, humidity, and light intensity) to form a time series feature vector. The hidden layer uses two LSTM layers (128 neurons per layer) to capture long- and short-term load fluctuations, such as weekday / weekend electricity usage differences and the impact of seasonal changes on heating and cooling loads. The output layer generates hourly load forecasts (electricity, heat, and cooling) for the next 24 hours, with a resolution of one hour.

[0098] Among them, the loss function takes the root mean square error (RMSE) as the optimization target, and the formula is:

[0099]

[0100] At the same time, the data set is divided into 80% for training and 20% for validation.

[0101] Step S3, constructing a multi-objective operation evaluation index system, includes the following steps:

[0102] Among them, energy consumption evaluation indicators include:

[0103] Total energy consumption of the park E total , expressed as:

[0104]

[0105] In the above, T is the number of time steps in the scheduling period (scheduling period is 24 hours, time step is 1 hour), E gen,i (t) is the power generation of the i-th power generation equipment at time t, E grid,j (t) is the amount of electricity purchased by the jth grid interaction point at time t, E storage,k (t) is the discharge capacity of the kth energy storage device at time t;

[0106] Energy consumption per unit area E area , expressed as:

[0107]

[0108] In the above, S is the total area of ​​the park.

[0109] Among them, environmental protection evaluation indicators include:

[0110] Carbon emissions total , expressed as:

[0111]

[0112] Above, C i (t) is the carbon emission of the i-th power generation equipment at time t, C grid,j(t) is the carbon emissions from electricity purchases at the jth grid interaction point at time t.

[0113] Among them, the carbon emission calculation of the i-th power generation equipment is expressed as:

[0114] C i (t) = λ i ×F i (t);

[0115] Above, λ i is the carbon emission coefficient of the i-th power generation equipment, F i (t) is the fuel consumption of the i-th power generation equipment at time t.

[0116] Among them, the power supply reliability evaluation indicators include:

[0117] Among them, the system power outage time T outage Calculation, expressed as:

[0118]

[0119] Above, T out (t) is the power outage time at time t, and T out (t)=0, otherwise it is the power outage duration.

[0120] Among them, the number of power outages N outage , expressed as:

[0121]

[0122] Among them, the cooling reliability evaluation indicators include:

[0123] Among them, the cooling temperature stability ΔT cool , expressed as:

[0124]

[0125] Above, T cool (t) is the actual cooling temperature at time t, T set_cool To set the cooling temperature.

[0126] Among them, the number of cooling interruptions N cool_outage , expressed as:

[0127]

[0128] Above, Q cool (t) is the cooling power at time t.

[0129] The optimization dispatch model for the day-ahead economic dispatch phase is constructed, including minimizing operating costs as the core goal, integrating fuel costs, grid interaction costs, equipment maintenance costs, unit start-up and shutdown costs, and revenue from heat and cooling sales, and constructing an objective function, which is expressed as:

[0130]

[0131] Among them, fuel cost FC i (t), expressed as:

[0132]

[0133] Above, P i (t) is the power generated by the i-th power generation equipment at time t, F i (P i (t)) is the fuel consumption function of the power generation equipment, η i (P i (t)) is the efficiency function of power generation equipment, C fuel,i For fuel prices.

[0134] Among them, the grid interaction fee GIC j (t), expressed as:

[0135] GIC j (t) = P grid_buy,j (t)×C buy (t)-P grid_sell,j (t)×C sell (t);

[0136] Above, P grid_buy,j (t) is the power purchased by the jth grid interaction point at time t, P grid_sell,j (t) is the electricity sold, C buy (t) and C sell (t) are the purchase and sales prices of electricity, respectively.

[0137] Among them, equipment maintenance cost MC k (t), expressed as:

[0138] MC k (t) = P k (t)×τ k ×C maintain,k ;

[0139] Above, P k (t) is the power of the kth device at time t, τk is the device operation time, C maintain,k is the maintenance cost coefficient.

[0140] Among them, the unit start-up and shutdown costs SSC p (t), expressed as:

[0141] SSC p (t)=Δu p (t)×C start,p ;

[0142] Above, Δu p (t)=|u p (t)-u p (t-1)|,u p (t) is the unit start-stop state variable, C start,p The heat sales revenue HSR is the startup cost. r (t), expressed as:

[0143] CSR s (t) = Q cool_sell,s (t)×C cool_sell (t);

[0144] Above, Q cool_sell,s (t) is the heat purchased by the sth heat user at time t, C cool_sell (t) is the heat price. Among them, the cold sales revenue CSR s (t), expressed as:

[0145] CSR s (t) = Q cool_sell,s (t)×C cool_sell (t);

[0146] Above, Q cool_sell,s (t) is the cooling capacity purchased by the sth cooling user at time t, C cool_sell (t) is the cooling price. The calibration constraints include: power balance constraints, equipment operation constraints, and environmental protection constraints; among them, power balance constraints include:

[0147] Power balance:

[0148]

[0149] Thermal balance:

[0150]

[0151] Cold force balance:

[0152]

[0153] Equipment operation constraints include:

[0154] Power generation equipment output constraints:

[0155]

[0156] Energy storage equipment constraints:

[0157]

[0158] SOC min ≤SOC(t)≤SOC max ;

[0159] 0≤P storage,in (t)≤P in_max ;

[0160] 0≤P storage,out (t)≤P out_max ;

[0161] Unit start and stop constraints:

[0162]

[0163] Environmental protection constraints include:

[0164] Carbon emission constraints:

[0165]

[0166] Pollutant emission constraints:

[0167]

[0168] Solve the model and develop a comprehensive energy optimization scheduling plan. This involves inputting the constructed multi-objective optimization scheduling model into a hybrid linear integer programming solver such as CPLEX or Gurobi. Using a branch-and-bound algorithm, the optimal solution for the decision variables for energy storage device charge and discharge status, flexible load regulation, and the start and stop and output status of the controllable cogeneration units is obtained, while satisfying all constraints. This results in a comprehensive energy optimization scheduling plan for the next 24 hours.

[0169] Specifically, in the application, the “Binhai Smart Industrial Park” covers an area of ​​80,000m 2 It includes electronic manufacturing plants (4 buildings), cold chain storage centers (2 buildings), and office buildings (1 building), with a peak power load of 1200kW, a heating load of 1500kW in winter, and a cooling load of 1800kW in summer.

[0170] The energy system configuration is as follows:

[0171] Distributed power generation: solar photovoltaic (installed capacity 800kWp, average daily power generation 3200kWh), gas turbines (2 units, each with a power of 500kW, fueled by natural gas, carbon emission coefficient 0.2kg / kWh).

[0172] Energy storage system: lithium battery energy storage (capacity 1.5MWh, SOC range 20%-90%), ice storage system (cold storage capacity 2000kWh).

[0173] Energy interaction: linked with the peak and valley electricity prices of the power grid (peak period 8:00-22:00, 1.2 yuan / kWh; valley period 22:00-8:00, 0.3 yuan / kWh), and supports the access of surplus electricity to the grid (0.5 yuan / kWh).

[0174] In addition, taking a working day in winter as an example, real-time data collection is shown in Table 1:

[0175] Table 1 Data collection information table

[0176]

[0177]

[0178] Among them, the LSTM load forecast results for the next 24 hours are shown in Table 2:

[0179] Table 2 Prediction data table

[0180]

[0181] In addition, the scheduling method of the present invention generates a day-ahead optimized scheduling plan, which is as follows:

[0182] This strategy minimizes operating costs (including fuel, grid connection fees, and equipment maintenance costs) while balancing carbon emissions constraints (a daily limit of 120 tons) and heat supply reliability. The optimization results in a total operating cost of 18,200 yuan / day, a 22% reduction compared to traditional scheduling. The time-of-day scheduling strategy is generated as follows:

[0183] During the off-peak hours (midnight to 8:00 AM), when the PV system is not operating, the grid purchases 300 kWh of electricity (at 0.3 yuan / kWh) during the off-peak period, prioritizing charging the lithium-ion battery (SOC increases from 20% to 70%). The ice storage system utilizes off-peak electricity to produce ice (increasing cooling capacity by 300 kWh), while heat is provided by a gas turbine cogeneration system (generating 200 kW of electricity and 800 kW of heat).

[0184] From midnight to 12:00 (the start of the peak period), the photovoltaic system activates (generating 300-600kW), supplying the factory's power load (600-800kW). The shortfall is met by lithium battery discharge (-80kW) and electricity purchased from the grid (peak electricity price of 1.2 yuan / kWh, purchasing 100kW of electricity). Heating is provided entirely by the gas turbine (generating 300kW and providing 1000kW of heat), meeting the needs of offices and factory buildings.

[0185] During the period from 12:00 PM to 4:00 PM (maximum photovoltaic generation), the photovoltaic system generates 600 kW of power, the lithium battery stops discharging (SOC = 50%), and the remaining power (600 kW + 200 kW from the gas turbine - 1100 kW load = -300 kW) requires 300 kW of electricity purchased from the grid (at a high peak price). The gas turbine operating mode is adjusted: power generation is reduced to 150 kW, heating power is increased to 1500 kW (to meet peak heat load), and fuel consumption is increased to 350 L / h.

[0186] During the evening peak hours of 6:00 PM to 10:00 PM, the photovoltaic system is shut down, the lithium battery is fully discharged (-150 kW, SOC reduced to 20%), and the gas turbine generates 500 kW of power at full capacity. However, 450 kW of electricity is still purchased from the grid (at the peak electricity price, accounting for 35% of the cost). The cooling load is relieved by the ice storage system (300 kWh / h), and the shortfall is met by the electric chiller (200 kW).

[0187] During the flat period from 10:00 PM to 12:00 AM, when the grid electricity price falls back to 0.6 yuan / kWh, the gas turbine is shut down and the lithium battery is charged to a SOC of 30%. Heat is then supplied by the heat storage tank (remaining 200 kWh), with the shortfall being supplemented by the waste heat recovery system (industrial waste heat utilization rate increased to 40%).

[0188] Constraints include: Power balance: The error between electricity and heat supply and demand is less than 1.5% during each period, and the error between cooling load supply and demand is less than 2%. Energy storage constraints: The lithium battery SOC is always between 20% and 90%, and the ice storage capacity is maintained between 50% and 90%. Environmental constraints: Total daily carbon emissions are 112 tons, 8 tons below the limit. Reliability: Heating temperature fluctuations are controlled within 20±0.5°C, with no power interruptions.

[0189] In summary, with the help of the above technical solution of the present invention, the following effects can be achieved:

[0190] 1. The present invention realizes multi-objective collaborative optimization. By constructing a multi-objective operation evaluation index system and an optimization scheduling model, it breaks through the limitations of traditional single objectives, realizes the collaborative optimization of energy consumption, environmental protection and reliability, and improves the comprehensiveness and scientificity of smart park energy management.

[0191] 2. The present invention improves energy utilization efficiency by coordinating the scheduling of multiple energy devices, fully tapping the potential of renewable energy, rationally allocating energy storage equipment and flexible loads, reducing energy loss, significantly improving the energy utilization efficiency of the park, and effectively reducing total energy consumption and energy consumption per unit area. At the same time, environmental constraints prompt the park to give priority to the use of clean energy and high-efficiency environmentally friendly equipment, significantly reducing carbon emissions and pollutant emissions, helping the park achieve green and low-carbon development goals, enhancing the park's environmental competitiveness, and further achieving reliable energy supply guarantees. Based on the reliability evaluation indicators and constraints of power supply, heating, and cooling, the optimized scheduling plan can effectively reduce the number and duration of energy supply interruptions, ensure the stable operation of production and life of park users, and improve the quality of park energy services.

[0192] 3. The present invention achieves strong adaptability and versatility. By relying on real-time data collection and dynamic model construction, it can quickly respond to changes in park energy demand and equipment status. It is suitable for smart parks of different scales and energy structures and has broad promotion and application value.

[0193] The foregoing is merely a preferred embodiment of the present invention and is not intended to limit the present invention. A person skilled in the art will readily appreciate other embodiments of the present invention after considering the disclosure in the specification and examples. This application is intended to cover any variations, uses, or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered merely exemplary, and the true scope and spirit of the present invention are indicated by the claims.

[0194] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A comprehensive energy optimization scheduling method for a smart park, characterized by: The following steps are involved: Conduct real-time data collection from multiple sources in advance, and obtain operating parameters of photovoltaic, wind power generation equipment, gas turbines, and energy storage systems, as well as user-side electricity, heat, and cooling load data in real time through a distributed sensor network; Build a load forecasting model based on the LSTM neural network, input historical load data and meteorological data, and output hourly load forecast values ​​for the next 24 hours; Construct multi-objective operation evaluation indicators, including at least energy consumption evaluation indicators, environmental protection evaluation indicators, power supply reliability evaluation indicators and cooling supply reliability evaluation indicators; An optimization scheduling model for the day-ahead economic scheduling stage is constructed with the core goal of minimizing operating costs. The objective function is constructed by integrating fuel costs and grid interaction costs. Power balance, equipment operation, and environmental protection are used as constraints to obtain the optimal scheduling solution through a hybrid linear integer programming solver.

2. The smart park comprehensive energy optimization scheduling method according to claim 1 is characterized in that: The energy consumption evaluation indicators include: total energy consumption of the park E total , expressed as: Among them, T is the number of time steps in the scheduling period, E gen,i (t) is the power generation of the i-th power generation equipment at time t, E grid,j (t) is the amount of electricity purchased by the jth grid interaction point at time t, E storage,k (t) is the discharge capacity of the kth energy storage device at time t; Energy consumption per unit area E area , expressed as: Where S is the total area of ​​the park.

3. The smart park comprehensive energy optimization scheduling method according to claim 2 is characterized in that: The environmental evaluation indicators include: carbon emissions C total , expressed as: Above, C i (t) is the carbon emission of the i-th power generation equipment at time t, C grid,j (t) is the carbon emissions of electricity purchased by the jth grid interaction point at time t; Among them, the carbon emission calculation of the i-th power generation equipment is expressed as: C i (t)=λ i ×F i (t); Above, λ i is the carbon emission coefficient of the i-th power generation equipment, F i (t) is the fuel consumption of the i-th power generation equipment at time t.

4. The smart park comprehensive energy optimization scheduling method according to claim 1 is characterized in that: The power supply reliability evaluation index includes: system power outage time T outage Calculation, expressed as: Above, T out (t) is the power outage time at time t, and T out (t) = 0, otherwise it is the duration of the power outage; Among them, the number of power outages N outage , expressed as:

5. The smart park comprehensive energy optimization scheduling method according to claim 4 is characterized in that: The cooling reliability evaluation index includes: cooling temperature stability ΔT cool , expressed as: Above, T cool (t) is the actual cooling temperature at time t, T set_cool To set the cooling temperature; Among them, the number of cooling interruptions N cool_outage , expressed as: Above, Q cool (t) is the cooling power at time t.

6. The smart park comprehensive energy optimization scheduling method according to claim 5 is characterized in that: The objective function of constructing the optimal dispatch model in the day-ahead economic dispatch stage is expressed as: Among them, fuel cost FC i (t), expressed as: Above, P i (t) is the power generated by the i-th power generation equipment at time t, F i (P i (t)) is the fuel consumption function of the power generation equipment, η i (P i (t)) is the efficiency function of power generation equipment, C fuel,i for fuel prices; Among them, the grid interaction fee GIC j (t), expressed as: GIC j (t)=P grid_buy,j (t)×C buy (t)-P grid_sell,j (t)×C sell (t); Above, P grid_buy,j (t) is the power purchased by the jth grid interaction point at time t, P grid_sell,j (t) is the electricity sold, C buy (t) and C sell (t) The electricity purchase and sales prices respectively; Among them, equipment maintenance cost MC k (t), expressed as: MC k (t)=P k (t)×τ k ×C maintain,k ; Above, P k (t) is the power of the kth device at time t, τk is the device operation time, C maintain,k is the maintenance cost coefficient; Among them, the unit start-up and shutdown costs SSC p (t), expressed as: SSC p (t)=Δu p (t)×C start,p ; Above, Δu p (t)=|u p (t)-u p (t-1)|,u p (t) is the unit start-stop state variable, C start,p for startup costs; Among them, heat sales revenue HSR r (t), expressed as: CSR s (t)=Q cool_sell,s (t)×C cool_sell (t); Above, Q cool_sell,s (t) is the heat purchased by the sth heat user at time t, C cool_sell (t) is the heat price; Among them, sales of cold revenue CSR s (t), expressed as: CSR s (t)=Q cool_sell,s (t)×C cool_sell (t); Above, Q cool_sell,s (t) is the cooling capacity purchased by the sth cooling user at time t, C cool_sell (t) is the cold price.

7. The smart park comprehensive energy optimization scheduling method according to claim 6 is characterized in that: The construction of the optimization scheduling model for the day-ahead economic scheduling stage also includes: calibrating constraint conditions, including power balance constraints, equipment operation constraints, and environmental protection constraints; Among them, the power balance constraints include: Power balance: Thermal balance: Cold force balance: Equipment operation constraints include: Power generation equipment output constraints: Energy storage equipment constraints: SOC min ≤SOC(t)≤SOC max ; 0≤P storage,in (t)≤P in_max ; 0≤P storage,out (t)≤P out_max ; Unit start and stop constraints: Environmental protection constraints include: Carbon emission constraints: Pollutant emission constraints:

8. The smart park comprehensive energy optimization scheduling method according to claim 1 is characterized in that: The hybrid linear integer programming solver obtains the optimal scheduling plan, including: using CPLEX or Gurobi hybrid linear integer programming solver, solving through a branch and bound algorithm, outputting the optimal decision variables for the charging and discharging status of the energy storage device, the flexible load adjustment amount, and the start and stop and output status of the unit, to form a comprehensive energy optimization scheduling plan for the next 24 hours.

Images