Multi-time-scale park integrated energy system distribution robust optimization scheduling method

Abstract

The invention discloses a multi-time-scale park integrated energy system distribution robust optimization scheduling method, which comprises the following steps of: constructing an electric heating collaborative system model taking a combined heat and power generation unit as a core, and introducing a carbon transaction mechanism; constructing a confidence set in combination with a 1-norm and an infinity-norm, and respectively making a robust start-stop plan and a flexible operation strategy in day-ahead and intra-day two-stage scheduling; a column and constraint generation algorithm is adopted to decompose the constructed day-ahead and intra-day two-stage model into a main problem and a sub-problem for repeated iterative solution, and an optimal scheduling scheme with both economical efficiency and robustness is obtained. According to the method, the uncertainty of new energy prediction is fully considered, and the coping of the park to the randomness of the new energy can be better played through processing of different time scales.

Description

Technical Field

[0001] This invention relates to the field of integrated energy system regulation technology in industrial parks, specifically a multi-timescale integrated energy system optimization scheduling method for industrial parks. Background Technology

[0002] With the rapid development of new energy sources, energy consumption has become a significant challenge facing humanity. Traditional energy systems are increasingly revealing problems such as low resource utilization and severe environmental pollution, making it difficult to meet the ever-growing energy demands. Park Integrated Energy Systems (PIES), through coordinated interaction and multi-energy complementarity among source-grid-load-storage systems, integrate various energy resources within a park, including electricity, heat, and natural gas, and have become a hot research topic in energy systems.

[0003] Traditionally, power and heating systems have been planned and operated independently, lacking a coordination mechanism, resulting in low energy efficiency. However, in actual operation, the coupling between the power and heating systems is constantly deepening, with significant mutual influence and constraints between them. Among numerous coupled devices, combined heat and power (CHP) units have attracted much attention due to their high efficiency, cleanliness, and low carbon emissions. Currently, many studies have explored the economic dispatch problem of integrated power-heat systems including CHP units using adaptive real-valued coding genetic algorithms and bee colony optimization algorithms. However, CHP units exhibit strong electrothermal coupling characteristics, making their power output easily limited by heat load and limiting their regulation capacity. Therefore, thermal storage devices or electric boilers are usually required to expand the output range and enhance regulation flexibility.

[0004] Currently, stochastic programming and robust optimization methods are mainly used to address uncertainties in PIES (Personalized Injection Systems). However, stochastic programming has high computational complexity when generating a large number of samples; while robust optimization, although providing guarantees in extreme scenarios, often leads to overly conservative scheduling strategies, affecting economic efficiency. Therefore, Distributed Robust Optimization (DRO) has been proposed as a more adaptive optimization method to solve source-load uncertainty problems. DRO finds the most unfavorable probability distribution within a known confidence set, thus achieving the optimization objective without relying on a large amount of historical data, while overcoming the drawback of overly conservative results. By introducing a comprehensive norm to constrain the probability distribution of the scenario, an effective balance is achieved between system economy and robustness. However, existing DRO algorithm research mainly focuses on the day-ahead scheduling stage and has not fully considered the problem of prediction accuracy variations across multiple time scales. Since the prediction accuracy of new energy sources increases with shortening time scales, exploring its application in real-time scheduling has significant practical implications.

[0005] Patent application CN116780649A discloses a distributed robust optimization operation method for multi-energy complementary utilization. This patent constructs an uncertainty set driven by historical correlation based on similar daily data points, extracts the actual spatiotemporal correlation of renewable energy generation, and incorporates a polyhedral uncertainty set to minimize the occurrence of extreme uncertainty scenarios with very low probabilities. However, as mentioned above, this patent's distributed robust optimization focuses on the day-ahead scheduling stage and does not fully consider the variation in prediction accuracy across multiple time scales, thus failing to fully leverage the park's ability to cope with the stochasticity of new energy sources. Summary of the Invention

[0006] The technical problem to be solved by this invention is that traditional stochastic and robust optimization are conservative and limited in dealing with uncertainty, and have not fully considered the changes in prediction accuracy under multiple time scales, thus failing to fully leverage the park's ability to cope with the stochasticity of new energy.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A multi-timescale integrated energy system for industrial parks, comprising: Construct a power-heat synergy system model with combined heat and power units as the core, and introduce a carbon trading mechanism; By combining the 1-norm and the ∞-norm to construct a confidence set, robust start-stop plans and flexible operation strategies are formulated for day-ahead and intraday two-stage scheduling, respectively. The column and constraint generation algorithm is used to decompose the constructed two-stage model of day-ahead and day-intraday into the main problem and sub-problems, and iteratively solves them to obtain the optimal scheduling scheme that is both economical and robust.

[0008] In this embodiment, an electrothermal synergy system model is constructed with a combined heat and power (CHP) unit as the core, including the operation model of the CHP unit, the operation model of the gas boiler, the carbon capture system model, the operation model of the power-to-gas conversion equipment, and the operation model of the energy storage equipment.

[0009] In this embodiment, a confidence set is constructed by combining the 1-norm and the ∞-norm, and robust start-stop plans and flexible operation strategies are formulated for the day-ahead and intraday two-stage scheduling, respectively, including: The construction phase is based on operational targets measured in one-hour timescales, including the start-up and shutdown costs of each piece of equipment. The intraday phase is designed with operational targets on a 15-minute timescale, including the operating costs of each device and carbon trading costs. Set the constraints that enable the park's integrated energy system to operate under the day-ahead and intraday operational targets; Construct a system centered on the initial probability distribution value, containing the 1-norm and - The comprehensive norm of the norm is used as a constraint condition to constrain the probability distribution value of discrete new energy scenarios and obtain its feasible region.

[0010] In this embodiment, the operational target for the day-ahead phase, measured on an hourly timescale, is expressed by the following formula: (12); In the formula, The operational goals for the current phase are as follows: , Each of the units in the park Start-up and shutdown fees, This is a startup variable; a value of 1 indicates that all units in the park are powered on, otherwise it is 0. This is a shutdown variable; a value of 1 indicates that all units in the park are shut down, otherwise it is 0. The total number of time periods. This refers to the number of each unit in the park; During the intraday phase, the operational target, measured in 15-minute timeframes, is expressed by the following formula: (13); (14); In the formula, For the operating costs of equipment in the park, The probability value of scenarios that contribute to new energy sources. This represents the total number of discrete scenes. For each unit in the park The unit's operating and maintenance costs, for Each unit in the park Operating power The penalty coefficient for abandoned wind turbines and photovoltaic power generation in the industrial park. This represents the upper limit of the output power of photovoltaic and wind turbines. , For photovoltaics, For the fan, for The operating power of photovoltaic and wind turbines at all times. For the carbon trading costs of the park, Price per unit of carbon emissions This is the initial carbon allowance for the park. This represents the actual carbon emissions of the park.

[0011] In this embodiment, the constraints that enable the integrated energy system of the park to operate under the day-ahead and intraday operational targets include: Current unit start-up and shutdown constraints: (17); In the formula, This is a startup variable. This is a shutdown variable; Intraday operational balance constraints: (18); ,(19) (20); ,(twenty one); In the formula, , They are respectively The output power of photovoltaic and wind turbines at all times This refers to the power generation capacity of the gas turbine. , They are respectively The charging and discharging power of the battery at all times. The electrical energy consumed by the electro-gas conversion equipment. This represents the total energy consumption of the carbon capture system. , These are the park's electrical and thermal load values, respectively. For the exchange of electricity with generating units in the upper-level power grid, This refers to the heating capacity of the waste heat boiler. Gas-fired boiler Real-time carbon emissions , They are respectively The heat storage and release power of the heat storage tank at all times. for The state of charge of the battery at all times. , These are the lower and upper limits of energy storage capacity. for The thermal energy stored in the heat storage tank at all times , These are the lower and upper limits of the thermal storage tank capacity. , These are the lower and upper limits of the energy storage charging and discharging capacity. , These are the lower and upper limits of the heat storage tank's charging and discharging capacity.

[0012] In this embodiment, the feasible region is: ,(twenty two); In the formula, For feasible regions, The probability value of scenarios that contribute to new energy sources. This represents the total number of discrete scenes. The first one obtained from actual operating data Initial probability values ​​for a discrete scene. , In the 1-norm, respectively -Probability tolerance limit under norm constraints The constraint is a 1-norm constraint. for - Norm constraint.

[0013] In this embodiment, at the 1-norm, -Probability tolerance limit under norm constraint , It can be obtained through the following formula: (25); In the formula, For the number of historical scenes, , These are the 1-norm of the probability distribution values ​​and - Confidence level of norm It is a logarithmic function.

[0014] In this embodiment, obtaining an optimal scheduling scheme that combines economy and robustness includes: The main problem is to find the optimal solution that satisfies the constraints under the known adverse probability distribution, to carry out the day-ahead scheduling of the park's integrated energy system, and to provide a lower bound for the overall operation target of the park. Given the variables in the main problem, find the worst-case probability distribution under real-time operation and return it to the main problem for use in the next iteration, providing an upper bound for the overall operation goal of the park.

[0015] In this embodiment, the main problem is to find the optimal solution that satisfies the constraints under a known adverse probability distribution, to perform day-ahead scheduling of the park's integrated energy system, and to provide a lower bound for the overall operational objectives of the park, including: ,(26; (27); In the formula, For the robust decision variables in the first stage, This is the first stage of the variable set. For the first Second-stage variables in each scenario For the second-stage variables in the new energy forecasting scenario. Forecast power output value for new energy sources This is the upper limit of the second phase target. For the second stage of the variable set, This is the set of constants related to the predicted scenario for the first-stage objective. For the number of iterations, To predict the day-ahead scheduling target in the scenario, , These are the predicted scenarios for the second phase of intraday operation and carbon trading targets. , These refer to the second phase of intraday operation under discrete scenarios and carbon trading targets. For the first The probability value of new energy power output scenarios in the next iteration. This represents the total number of discrete scenes. For the first New energy output values ​​under discrete scenarios.

[0016] In this embodiment, given the variables of the main problem, the worst-case probability distribution under intraday operation is found and returned to the main problem for use in the next iteration, providing an upper bound value for the overall operational goal of the park: (29); In the formula, Solve the objective for the subproblems. For the first Park operation objectives under discrete scenarios For feasible regions, This is the optimal set of variables for the first stage in the prediction scenario.

[0017] Compared with the prior art, the beneficial effects of the present invention are: This invention constructs a coordinated optimization scheduling model for an integrated power-heat system centered on a combined heat and power (CHP) unit. The objective function encompasses the unit's power generation cost, wind and solar curtailment cost, and carbon trading cost. Addressing the conservatism and limitations of traditional stochastic and robust optimization methods in handling uncertainty, this invention fully considers the uncertainty of renewable energy prediction and proposes an improved data-driven two-stage partial Bruker optimization scheduling model. Combining multi-timescale characteristics, different timescales are used to design scheduling strategies in the first and second stages, specifically, robust start-up and shutdown plans and flexible operation strategies are formulated in the day-ahead (1-hour timescale) and intraday (15-minute timescale) scheduling stages, respectively. Finally, numerical examples verify the effectiveness of the proposed model in improving renewable energy absorption capacity, demonstrating the feasibility and superiority of the proposed partial Bruker optimization method in power-heat system scheduling.

[0018] This invention is based on an improved split-bulk algorithm. It processes the two-stage algorithm at different time scales on the original basis, which can better enable the park to cope with the randomness of new energy.

[0019] This invention constructs an optimization model based on master and sub-problems, and uses the CCG algorithm to perform master-slave iterative solution on the model to obtain an optimal scheduling scheme that is both economical and robust. Attached Figure Description

[0020] Figure 1 This is a flowchart of a multi-timescale integrated energy system in a park with distributed bar optimization scheduling method, as an embodiment of the present invention.

[0021] Figure 2 This is a structural diagram of the integrated energy system in the park according to an embodiment of the present invention.

[0022] Figure 3 This is a schematic diagram of a discrete scenario for wind power generation in a park integrated energy system according to an embodiment of the present invention.

[0023] Figure 4 This is a schematic diagram of a discrete scenario of photovoltaic power generation in a park integrated energy system according to an embodiment of the present invention.

[0024] Figure 5 This is a schematic diagram of the first stage of device startup in an embodiment of the present invention.

[0025] Figure 6 This is a schematic diagram of the second stage of equipment scheduling (electrical energy) in an embodiment of the present invention.

[0026] Figure 7 This is a schematic diagram of the second-stage equipment scheduling action (thermal energy) according to an embodiment of the present invention.

[0027] Figure 8 This is a schematic diagram comparing the demand response results of an embodiment of the present invention.

[0028] Figure 9 This is a schematic diagram comparing the results of new energy reduction in an embodiment of the present invention. Detailed Implementation

[0029] To facilitate understanding of the technical solution of the present invention by those skilled in the art, the technical solution of the present invention will now be further described in conjunction with the accompanying drawings.

[0030] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0031] Please see Figure 1 , 2 As shown, this invention provides a multi-timescale integrated energy system in a park with distributed bar optimization scheduling method, including: S10 constructs a combined heat and power (CHP) system model with CHP units at its core and introduces a carbon trading mechanism.

[0032] In one embodiment of the present invention, an electrothermal synergy system model is constructed with a combined heat and power (CHP) unit as the core, including an operation model of the CHP unit, an operation model of the gas-fired boiler, a carbon capture system model, an operation model of the power-to-gas (EPG) equipment, and an operation model of the energy storage equipment. Specifically, it includes the following steps.

[0033] S11, Obtaining the operating model of the combined heat and power (CHP) unit includes: The CHP unit comprises a gas turbine (GT) and a waste heat boiler (WHB). The GT generates electricity by burning natural gas, and the waste heat produced is recovered by the WHB and used for heating. The relationship between the electrical power output of the GT and the gas consumption can be expressed as: ,(1); (2); In the formula, This indicates the time scale for optimizing the park's scheduling; This indicates the electrical energy resources within the park; This indicates the thermal energy source in the park; This represents the actual carbon emissions of the CHP unit; Power generation for GT; This refers to the heating power of the WHB. , , Parameters for calculating carbon emissions from gas-fired power units; For GT's power generation efficiency; It has a low calorific value for natural gas; Natural gas consumed by GT; Waste heat from GT exhaust; This refers to the heat loss rate; For waste heat recovery efficiency; The flue gas recovery rate of WHB.

[0034] S12, Construct the operating model of the gas boiler (GB). The GB provides heat by burning natural gas. The operating model is shown below: (3); (4); In the formula, GB Real-time carbon emissions; The heating power is GB. The heating efficiency is in GB. It has a low calorific value for natural gas; Natural gas consumed by GB. , , These are the calculation parameters for the gas-fired boiler operation model.

[0035] S13. Construct a carbon capture system (CCS) model to capture CO2 in the park's integrated energy system, separating and concentrating carbon dioxide from the CO2-containing gas stream, and then supplying it to an electricity-to-gas converter to convert it into natural gas. The operating model is as follows: (5); (6); (7); In the formula, For CCS Real-time carbon capture volume; For CHP units Real-time carbon emissions; GB Real-time carbon emissions; For carbon capture efficiency; Total energy consumption of CCS; Fixed energy consumption for CCS; Energy consumption for CCS operation; The energy required for CCS to process one unit of CO2.

[0036] S14. Construct an operating model for a power-to-gas (P2G) system. The P2G system generates natural gas by electrolyzing water and reacting it with CO2 gas captured and transported by the CCS. The operating model is shown below: (8); (9); In the formula, Natural gas generated for P2G; The heat energy that can be converted from a unit of electrical energy; For P2G working efficiency; The electrical energy consumed by P2G; The amount of CO2 consumed by P2G; The amount of CO2 required to generate one unit of natural gas power. It has a low calorific value.

[0037] S15, Construct the operational model of the energy storage equipment. The energy storage equipment mainly includes batteries (BT) and heat storage tanks (HST), and the operational model is shown below: (10); ,(11) In the formula: for The state of charge of BT at any given time; , for The charging and discharging power of BT at any given time, in kW; , for The charging and discharging efficiency of BT at any given time; for The thermal energy stored in HST at any given time; , for The heat storage and release power of HST at any given time; HST energy loss rate; , The heat storage and release efficiency of HST.

[0038] S20, combining the 1-norm and ∞-norm to construct a confidence set, and formulating robust start-stop plans and flexible operation strategies in the day-ahead and intraday two-stage scheduling to improve system flexibility and predictive adaptability, specifically including the following steps: S21, the day-ahead phase, establishes operational targets on an hourly timescale, including start-up and shutdown costs for each piece of equipment: (12); In the formula: The operational goals for the current phase are as follows: , Each of the units in the park Start-up and shutdown fees, This is a startup variable; a value of 1 indicates that all units in the park are powered on, otherwise it is 0. This is a shutdown variable; a value of 1 indicates that all units in the park are shut down, otherwise it is 0. The total number of time periods. This refers to the number of units in each part of the park.

[0039] S22 establishes an intraday phase, with operational targets measured in 15-minute timeframes, and includes the operating costs of each device and carbon trading costs. (13); (14); In the formula, For the operating costs of equipment in the park, The probability value of scenarios that contribute to new energy sources. This represents the total number of discrete scenes. For each unit in the park The unit's operating and maintenance costs, for Each unit in the park Operating power The penalty coefficient for abandoned wind turbines and photovoltaic power generation in the industrial park. This represents the upper limit of the output power of photovoltaic and wind turbines. , For photovoltaics, For the fan, for The operating power of photovoltaic and wind turbines at all times. For the carbon trading costs of the park, Price per unit of carbon emissions This is the initial carbon allowance for the park. This refers to the actual carbon emissions of the industrial park. The various units within the park mentioned here include, as described above, cogeneration units, gas-fired boilers, carbon capture systems, power-to-gas conversion equipment, energy storage equipment, and conventional thermal power units connected to the upstream power grid.

[0040] In this embodiment, the initial carbon allowance and actual carbon emissions of the park are obtained using the following methods: (15); ,(16) In the formula, , Carbon quota coefficient for unit power and heat output. This refers to the electrical power exchanged with generating units in the upstream power grid. This represents the total power of the CHP unit.

[0041] S23, Set the constraints for the operation of the park's integrated energy system under the day-ahead and intraday operational targets: Current unit start-up and shutdown constraints: (17); In the formula, This is a startup variable. This is a shutdown variable; Intraday operational balance constraints: (18); ,(19) (20); ,(twenty one); In the formula, , They are respectively The output power of photovoltaic and wind turbines at all times This refers to the power generation capacity of the gas turbine. , They are respectively The charging and discharging power of the battery at all times. The electrical energy consumed by the electro-gas conversion equipment. This represents the total energy consumption of the carbon capture system. , These are the park's electrical and thermal load values, respectively. For the exchange of electricity with generating units in the upper-level power grid, This refers to the heating capacity of the waste heat boiler. Gas-fired boiler Real-time carbon emissions , They are respectively The heat storage and release power of the heat storage tank at all times. for The state of charge of the battery at all times. , These are the lower and upper limits of energy storage capacity. for The thermal energy stored in the heat storage tank at all times , These are the lower and upper limits of the thermal storage tank capacity. , These are the lower and upper limits of the energy storage charging and discharging capacity. , These are the lower and upper limits of the heat storage tank's charging and discharging capacity.

[0042] S24, construct a system centered on the initial probability distribution value, containing the 1-norm and - The comprehensive norm of the norm is used to constrain the probability distribution values ​​of discrete new energy scenarios and obtain their feasible region: ,(twenty two); In the formula, For feasible regions, The probability value of scenarios that contribute to new energy sources. This represents the total number of discrete scenes. The first one obtained from actual operating data Initial probability values ​​for a discrete scene. , In the 1-norm, respectively -Probability tolerance limit under norm constraints The constraint is a 1-norm constraint. for - Norm constraint.

[0043] In this embodiment, The following confidence levels are satisfied: ,(twenty three); use and To express the confidence level of the probability distribution value, equation (24) can be transformed into equation (25): ,(twenty four); In the formula: , The sum of the 1-norms of the probability distribution values - Confidence level of the norm, in bold Let be the base of the natural logarithm function. The mathematical notation for probability, using the 1-norm as an example, represents all discrete scenarios. and The sum of the absolute values ​​of the differences is less than or equal to The probability needs to be greater than or equal to , This represents the total number of historical scenes.

[0044] From equations (23) and (24), we can obtain equation (25) as follows: (25); In the formula, It is a logarithmic function.

[0045] S30 employs the Column and Constraint Generation (CCG) algorithm to decompose the constructed two-stage model (day-to-day and intraday) into a master problem (MP) and subproblems (SP) for iterative solving to obtain an optimal scheduling scheme that combines economy and robustness. Specifically, this includes the following steps: S31: The main problem is to find the optimal solution that satisfies the constraints under the known adverse probability distribution, to carry out the day-ahead scheduling of the park's integrated energy system, and to provide a lower bound for the overall operation target of the park.

[0046] ,(26; (27); In the formula, For the robust decision variables in the first stage, This is the first stage of the variable set. For the first Second-stage variables in each scenario For the second-stage variables in the new energy forecasting scenario. Forecast power output value for new energy sources This is the upper limit of the second phase target. For the second stage of the variable set, This is the set of constants related to the predicted scenario for the first-stage objective. For the number of iterations, To predict the day-ahead scheduling target in the scenario, , These are the predicted scenarios for the second phase of intraday operation and carbon trading targets. , These refer to the second phase of intraday operation under discrete scenarios and carbon trading targets. For the first The probability value of new energy power output scenarios in the next iteration. This represents the total number of discrete scenes. For the first New energy output values ​​under discrete scenarios.

[0047] S32, given the variables in the main problem, finds the worst-case probability distribution under real-time operation and returns it to the main problem for use in the next iteration, providing an upper bound for the overall operational goals of the park: (28); In the inner min function of equation (28), each scenario problem is independent and can be processed using a parallel solution method, and is expressed as follows: Thus, equation (28) can be rewritten as: (29); In the formula, Solve the objective for the subproblems. For the first Park operation objectives under discrete scenarios For feasible regions, This is the optimal set of variables for the first stage in the prediction scenario. In summary, the two-stage model of equations (12) to (14) continuously iterates and updates variables between equations (26), (27), and (29) until a given accuracy value is reached, at which point the iteration stops.

[0048] To enable those skilled in the art to better understand the present invention, the numerical example analysis includes the following components: I. Case Description and Simulation Result Analysis Using a 1-hour scheduling interval within a day, the four discrete scenarios of new energy unit output within PIES are as follows: Figure 3 , Figure 4 As shown in Tables 1 and 2, the operating parameters of CHP units and other equipment and energy storage are used; the electricity price adopts the reference existing technology; the unit cost of wind and solar curtailment penalty is 0.03 yuan / kWh.

[0049] The start-up and shutdown cost of P2G equipment is 60 yuan per time; the initial energy storage capacity of BT is 90kW, and the charging and discharging efficiency is 0.96; the initial heat storage capacity of HST is 80kW, the charging and discharging efficiency is 0.98, and the energy self-loss rate is 0.025.

[0050] Table 1 Energy Equipment Parameters

[0051] Table 2 Energy Storage Equipment Parameters

[0052] In the first phase, the startup cost of each piece of equipment is used as the optimization objective. By pre-setting start-up and shutdown strategies in the first hour of the day, constraints are provided for equipment output scheduling on a 15-minute timescale throughout the day in the second phase, thereby reducing the intraday decision-making burden. The start-up and shutdown actions of each piece of equipment are as follows: Figure 5 As shown. Observation Figure 6 and Figure 7 It can be seen that from 00:00 to 07:00, the nighttime heat load is high, but the electrical load is underestimated. The GB and CHP units operate simultaneously to meet the heat demand, while the electrical load increases in the latter part of the night, with only the CHP unit outputting its full capacity to meet both heat and electricity demands. From 07:00 to 13:00, the heat load decreases. Due to the gradual decrease in electrical power during this short period, CHP and GB units alternately output power to meet the heat demand. From 13:00 to 22:00, the heat load gradually increases, while the electrical load is in a "low-high-low" state. Therefore, the CHP unit, as the main equipment, outputs its full capacity to meet both demands, while GB outputs a small amount to meet the heat demand. The HST releases stored heat, and all equipment operates in coordination. From 23:00 to 24:00, the electrical load is high, and CHP and GB continue to supply heat, with excess heat stored in the HST. From 23:00 to 24:00, both electrical and heat loads are low; WT is sufficient to meet the electrical load, and GB provides heat.

[0053] Due to the presence of P2G (Power-to-Generate) systems, during periods of abundant wind and solar resources (01:00-05:00 and 12:00-14:00), surplus wind and solar power output enables the electrolyzers to begin operating, absorbing a significant portion of the wind and solar power and preventing resource waste caused by a mismatch between load and power generation. Furthermore, during periods when heat load is significantly higher than electrical load (01:00-04:00), the CHP (Consumer Power Plant) units meet the heat load while generating surplus electricity, increasing the electrolyzer output power and, together with the CHP units, alleviating the conflict between electricity and heat demand to some extent.

[0054] Figure 8 To compare the demand response results, on the one hand, there is the characteristic of new energy sources to counteract peak shaving, and on the other hand, there is the influence of time-of-use electricity pricing and the substitution of heat load. In order to maximize the local absorption of new energy power and reduce the energy loss in the P2G electricity-to-gas process, the electricity load implements differentiated electricity load adjustment strategies during the periods of 0:00-5:00, 10:30-13:45, and 18:30-19:15, and the electricity load demand shows different degrees of growth.

[0055] Observing the heat load curve, it can be seen that it exhibits a clear characteristic of high demand at night and low demand during the day. Combined with the tiered carbon trading model, the park participates in the carbon trading market and is constrained by carbon trading costs. PIES chooses to schedule the heat load to transfer a portion of the heat load demand at night to the daytime period, thereby reducing the power consumption of gas equipment during the daytime period and reducing PIES's daily carbon emissions.

[0056] like Figure 9 As shown, compared to the case without source-load interaction, despite the addition of demand response, the amount of renewable energy curtailed increased by approximately 2083.3 kW in one day. While adding P2G equipment can partially consume surplus electricity, the power output of P2G units is limited and cannot fully absorb renewable energy generation. Taking discrete scenario 4 as an example, during periods of abundant wind resources (07:00–08:15), excess wind and solar power gradually increases the output of the electrolyzer. During periods of extremely abundant wind and solar resources (10:00–13:15), the electrolyzer operates at full capacity. Combined with the DR mechanism, almost all renewable energy output is effectively absorbed.

[0057] To verify the effectiveness of the data-driven multi-timescale DRO model, a comparative analysis was conducted by setting different historical sample sizes and confidence intervals. The specific results are shown in Tables 3 and 4. Table 3 shows that when the 1-norm probability confidence interval... , -Probability confidence interval of norm When M is 0.5 and 0.99 respectively, the total cost tends to decrease as M increases. As shown in the equation, increasing M lowers the probability tolerance limit, making the probability distribution value in extreme scenarios closer to the probability distribution value in the initial scenario. This reduces the overall robustness of the system, preventing the system from deploying more generators to mitigate the uncertainties of new energy sources and load, thus lowering the total cost. Furthermore, the calculation results show that the comprehensive norm constraint proposed in this invention, under the same number of historical data points, has a lower total cost compared to the other two constraint conditions, indicating that the model in this invention is less conservative.

[0058] As shown in Table 4, with , As the confidence interval increases, the total cost also increases. This is because a larger confidence interval increases the uncertainty within the feasible region, requiring more generating units to mitigate the uncertainties of new energy sources and loads, thus increasing the total cost. However, as... As the cost increases, the overall cost growth trend is relatively moderate, with an average growth rate of only 0.33%. In contrast, with... As the value increases, the overall cost growth trend becomes more pronounced, with an average growth rate of 0.56%. The above results indicate that when... , When the value is large, the 1-norm constraint has no effect on the confidence set, only... - Norms act as constraints.

[0059] Table 3 Comparison of results with different numbers of historical data

[0060] Table 4 Comparison of results at different confidence levels

[0061] This invention takes into account the uncertainty of wind and solar power output and establishes a two-stage, multi-timescale distributed robust optimization scheduling model for the comprehensive energy of the park. The conclusions are as follows.

[0062] 1. The calculation results of the data-driven, multi-timescale, two-stage sub-Bruker optimization algorithm proposed in this invention are related to the selected historical data and confidence intervals. Specifically, the more historical data selected, the lower the conservatism of the algorithm; the larger the confidence interval, the higher the conservatism of the algorithm.

[0063] 2. The multi-timescale split-Brow bar optimization algorithm combines multi-time and split-Brow bar optimization algorithms. It uses day-ahead decision-making to alleviate intraday decision-making pressure, while using intraday decision-making to resolve uncertainty. It has better accuracy for intraday scheduling than the ordinary split-Brow bar optimization algorithm.

[0064] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.

[0065] The above embodiments are merely examples of implementation methods of the invention. The scope of protection of the present invention is not limited to the above embodiments. For those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention.

Claims

1. A multi-timescale integrated energy system for a park, characterized in that, include: Construct a power-heat synergy system model with combined heat and power units as the core, and introduce a carbon trading mechanism; By combining the 1-norm and the ∞-norm to construct a confidence set, robust start-stop plans and flexible operation strategies are formulated for day-ahead and intraday two-stage scheduling, respectively. The column and constraint generation algorithm is used to decompose the constructed two-stage model of day-ahead and day-intraday into the main problem and sub-problems, and iteratively solves them to obtain the optimal scheduling scheme that is both economical and robust.

2. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 1, characterized in that, Construct a combined heat and power (CHP) system model with the CHP unit as the core, including the operation model of the CHP unit, the operation model of the gas boiler, the carbon capture system model, the operation model of the power-to-gas conversion equipment, and the operation model of the energy storage equipment.

3. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 1, characterized in that, By combining the 1-norm and ∞-norm to construct a confidence set, robust start-stop plans and flexible operation strategies are formulated for the day-ahead and intraday two-stage scheduling, including: The construction phase is based on operational targets measured in one-hour timescales, including the start-up and shutdown costs of each piece of equipment. The intraday phase is designed with operational targets on a 15-minute timescale, including the operating costs of each device and carbon trading costs. Set the constraints that enable the park's integrated energy system to operate under the day-ahead and intraday operational targets; Construct a system centered on the initial probability distribution value, containing the 1-norm and - The comprehensive norm of the norm is used as a constraint condition to constrain the probability distribution value of discrete new energy scenarios and obtain its feasible region.

4. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 3, characterized in that, The operational target for the current phase, measured in one-hour increments, is expressed by the following formula: ，（12）； In the formula, The operational goals for the current phase are as follows: , Each of the units in the park Start-up and shutdown fees, This is a startup variable; a value of 1 indicates that all units in the park are powered on, otherwise it is 0. This is a shutdown variable; a value of 1 indicates that all units in the park are shut down, otherwise it is 0. Total number of time periods This refers to the number of each unit in the park; During the intraday phase, the operational target, measured in 15-minute timeframes, is expressed by the following formula: ，（13）； ，（14）； In the formula, For the operating costs of equipment in the park, The probability value of scenarios that contribute to new energy sources. This represents the total number of discrete scenes. For each unit in the park The unit's operating and maintenance costs, for Each unit in the park Operating power The penalty coefficient for abandoned wind turbines and photovoltaic power generation in the industrial park. This represents the upper limit of the output power of photovoltaic and wind turbines. , For photovoltaics, For the fan, for The operating power of photovoltaic and wind turbines at all times. For the carbon trading costs of the park, Price per unit of carbon emissions This is the initial carbon allowance for the park. This represents the actual carbon emissions of the park.

5. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 3, characterized in that, The constraints that enable the integrated energy system of the park to operate under the day-ahead and intraday operational targets include: Recent unit start-up and shutdown constraints: ，（17）； In the formula, For boot variables, This is a shutdown variable; Intraday operational balance constraints: ，（18）； ，（19）； ，（20）； ，（21）； In the formula, , They are respectively The output power of photovoltaic and wind turbines at all times This refers to the power generation capacity of the gas turbine. , They are respectively The charging and discharging power of the battery at all times. The electrical energy consumed by the electro-gas conversion equipment. The total energy consumption of the carbon capture system. , These are the park's electrical and thermal load values, respectively. For the exchange of electricity with generating units in the upper-level power grid, This refers to the heating capacity of the waste heat boiler. Gas-fired boiler Real-time carbon emissions , They are respectively The heat storage and release power of the heat storage tank at all times. for The state of charge of the battery at all times. , These are the lower and upper limits of energy storage capacity. for The thermal energy stored in the heat storage tank at all times , These are the lower and upper limits of the thermal storage tank capacity. , These are the lower and upper limits of the energy storage charging and discharging capacity. , These are the lower and upper limits of the heat storage tank's charging and discharging capacity.

6. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 3, characterized in that, Its feasible region is: ，（22）； In the formula, For feasible regions, The probability value of scenarios that contribute to new energy sources. This represents the total number of discrete scenes. The first one obtained from actual operating data Initial probability values ​​for a discrete scene. , In the 1-norm, respectively -Probability tolerance limit under norm constraints The constraint is a 1-norm constraint. for - Norm constraint.

7. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 6, characterized in that, In the 1-norm, -Probability tolerance limit under norm constraint , It can be obtained through the following formula: ，（25）； In the formula, For the number of historical scenes, , These are the 1-norm of the probability distribution values ​​and - Confidence level of norm It is a logarithmic function.

8. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 1, characterized in that, To obtain an optimal scheduling scheme that is both economical and robust, including: The main problem is to find the optimal solution that satisfies the constraints under the known adverse probability distribution, to carry out the day-ahead scheduling of the park's integrated energy system, and to provide a lower bound for the overall operation target of the park. Given the variables in the main problem, find the worst-case probability distribution under real-time operation and return it to the main problem for use in the next iteration, providing an upper bound for the overall operation goal of the park.

9. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 8, characterized in that, The main problem involves finding the optimal solution that satisfies the constraints under a known adverse probability distribution, performing day-ahead scheduling of the park's integrated energy system, and providing lower bounds for the overall operational objectives of the park, including: ，（26）； ，（27）； In the formula, For the robust decision variables in the first stage, This is the set of variables for the first stage. For the first Second-stage variables in each scenario For the second-stage variables in the new energy forecasting scenario. Forecast power output for new energy sources This is the upper limit of the second phase target. For the second stage of the variable set, This is the set of constants related to the predicted scenario for the first-stage objective. For the number of iterations, To predict the day-ahead scheduling target in the scenario, , These are the predicted scenarios for the second phase of intraday operation and carbon trading targets. , These refer to the second phase of intraday operation under discrete scenarios and carbon trading targets. For the first The probability value of new energy power output scenarios in the next iteration. This represents the total number of discrete scenes. For the first New energy output values ​​under discrete scenarios.

10. The multi-timescale integrated energy system of a park with distributed bar optimization scheduling method according to claim 9, characterized in that, Given the variables in the main problem, find the worst-case probability distribution under real-time operation and return it to the main problem for use in the next iteration, providing an upper bound for the overall operational goals of the park: ，（29）； In the formula, Solve the objective for the subproblems. For the first Park operation objectives under discrete scenarios For feasible regions, This is the optimal set of variables for the first stage in the prediction scenario.

Images