Park integrated energy system scheduling method considering energy consumption of central air conditioner cluster

Abstract

The invention discloses a park integrated energy system scheduling method considering energy consumption of a central air conditioner cluster, and belongs to the technical field of energy management and optimization control. Generating a plurality of time sequence prediction samples in a future scheduling period; performing clustering analysis on each time sequence prediction sample, and selecting a plurality of typical day scene data; constructing a corresponding day-ahead scheduling optimization target based on the typical day scene data and the comfort degree deviation index; based on a day-ahead scheduling optimization target, solving through a day-ahead double-layer nested optimization model to obtain a scheduling result corresponding to each typical day scene; determining a next-day reference scheduling plan; and in the next-day park integrated energy system operation process, taking the next-day reference scheduling plan as an initial plan, and performing dynamic correction by adopting an intra-day rolling optimization and game coordination model. The method solves the problems that an existing park integrated energy system is insufficient in time scale coordination, the flexible utilization of a central air conditioner is limited, and the operation robustness is low.

Description

Technical Field

[0001] This invention belongs to the field of energy management and optimization control technology, and in particular relates to a method for scheduling a comprehensive energy system in a park that takes into account the energy consumption of a central air conditioning cluster. Background Technology

[0002] With the in-depth implementation of the "dual carbon" goals, energy systems are accelerating their evolution towards multi-energy complementarity and intelligence. Traditional single-energy systems, due to their single energy form, slow response, and low operating efficiency, can no longer meet the needs of integrated energy systems at the park level for high efficiency, cleanliness, and flexible dispatch.

[0003] The Comprehensive Integrated Energy System (CIES) integrates various energy sources such as electricity, heat, and cooling to achieve tiered energy supply and cascaded utilization, becoming an important form of future smart energy development. However, the operation of CIES is characterized by multi-timescale coupling, diverse load types, and high operational uncertainty, making its scheduling and optimization extremely challenging.

[0004] Research indicates that central air conditioning clusters typically account for over 40% of a park's total energy consumption, and their operational characteristics have a decisive impact on system peak-valley distribution, energy efficiency, and operational economy. Central air conditioning systems possess significant potential for flexible regulation; by rationally setting temperature, power, and start-stop strategies, peak shaving and valley filling, as well as energy efficiency improvements, can be effectively achieved. However, current research and practice still face the following major challenges: (1) The time scale is singular and lacks day-ahead and intraday coordination. Existing scheduling mostly adopts a phased independent strategy: day-ahead optimization focuses on long-term economics, while real-time scheduling focuses on short-term fluctuation response. The two lack coordination and it is difficult to take into account both global and local objectives. Especially when considering central air conditioning clusters, their delay characteristics are not utilized to achieve cross-time scale adjustment, and the system flexibility is limited.

[0005] (2) Insufficient flexible utilization of central air conditioning. Existing models mostly treat air conditioning as a rigid load, only performing power tracking or simple reduction control, ignoring its adjustability within the range of comfort allowable, resulting in it being in a "passive response" state in the system, which affects both energy utilization efficiency and user experience.

[0006] (3) Insufficient system optimization and robustness. Integrated energy systems involve multiple objectives (economy, low carbon emissions, comfort) and multiple constraints. Traditional algorithms converge slowly in high-dimensional scenarios and are prone to getting trapped in local optima. The fluctuation of renewable energy output and load forecasting errors further exacerbate scheduling risks.

[0007] Therefore, there is an urgent need for a collaborative scheduling method that can balance long-term planning and short-term response, fully tap the flexibility potential of central air conditioning, and possess strong robust optimization capabilities, so as to achieve efficient, low-carbon, and stable operation of the park's integrated energy system. Summary of the Invention

[0008] To address the aforementioned shortcomings in existing technologies, this invention provides a scheduling method for a park integrated energy system that considers the energy consumption of central air conditioning clusters. This method solves the problems of insufficient time-scale coordination, limited flexible utilization of central air conditioning, and low operational robustness in existing park integrated energy systems.

[0009] To achieve the above-mentioned objectives, the technical solution adopted by this invention is a method for scheduling a comprehensive energy system in a park that considers the energy consumption of a central air conditioning cluster, comprising: Establish a comprehensive energy system for the park, including power supply, cold and heat source equipment, energy storage units, and user-side air conditioning clusters; Construct a comfort deviation index for user-side air conditioning clusters; By selecting different historical load data, historical renewable energy output data, and historical meteorological data from the park's integrated energy system, and calling the trained deep learning-based load and renewable energy output prediction model, several time-series prediction samples for future scheduling cycles are generated; the time-series prediction samples include time-series data of electricity load, cooling load, and renewable energy output. Cluster analysis was performed on each time-series prediction sample to obtain several scene clusters; the time-series prediction sample closest to the center of each scene cluster was selected as the typical daily scene data of the corresponding scene cluster. Based on data from typical daily scenarios and comfort deviation indicators, daily scheduling optimization objectives corresponding to each typical daily scenario are constructed. Based on the corresponding daily scheduling optimization objectives, the scheduling results for each typical daily scenario are obtained by solving the problem through a daily double-nested optimization model. The load and renewable energy output prediction model trained based on the latest historical data is used to generate the next day's prediction time series. The distance between the next day's prediction time series and the time series prediction samples corresponding to each typical day scenario is measured, and the typical day scenario with the smallest distance is selected as the target scenario. The scheduling result corresponding to the target scenario is used as the baseline scheduling plan for the next day. During the operation of the park's integrated energy system the following day, the baseline scheduling plan for the following day is used as the initial plan, and dynamic correction is carried out using an intraday rolling optimization and game coordination model. The aforementioned day-ahead double-nested optimization model and intraday rolling optimization and game coordination model both use the improved leech swarm algorithm as the solver.

[0010] The beneficial effects of this invention are as follows: This invention employs a day-ahead and intraday double-layer nested mechanism, enabling the integrated energy system of the park to achieve optimal economic efficiency over a long timescale and possess real-time response and fluctuation compensation capabilities over a short timescale, effectively improving the coordination and stability of system scheduling. Through dual constraints of comfort temperature zones and power settings, the central air conditioning system is transformed from a rigid load into an adjustable flexible unit, which helps the integrated energy system of the park achieve peak shaving and valley filling and demand response, balancing comfort and energy efficiency. By comprehensively considering the interaction and conversion between electrical, thermal, and cooling energy flows, and combining rolling optimization and scenario correction mechanisms, the impact of renewable energy output fluctuations in the integrated energy system of the park on system scheduling performance is effectively reduced, improving the system's energy stability.

[0011] Furthermore, the power supply side, used for power supply, includes photovoltaic power generation units, wind power generation units, gas boilers, combined heat and power units, and the public power grid; The aforementioned heat and cold source equipment is used to realize the mutual conversion between electrical energy, thermal energy and cold energy, including electric chillers and lithium bromide absorption chillers; The energy storage unit is used to realize energy time migration and peak shaving and valley filling, including electrical energy storage system, thermal energy storage system and cold energy storage system.

[0012] The beneficial effects of the above-mentioned further scheme are as follows: by clarifying the four-layer structure of the park's integrated energy system, the coordinated supply of multiple energy sources such as electricity, heat, and cooling is realized, providing a complete physical basis for the subsequent establishment of an energy coupling model and optimized scheduling.

[0013] Furthermore, the operational constraints of the integrated energy system in the park include energy balance constraints, energy equipment operation constraints, and energy storage unit dynamic constraints:

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

[0020]

[0021]

[0022]

[0023]

[0024] in, Photovoltaic power generation unit Power generation at any given moment; Wind power generation unit Power generation at any given moment; For combined heat and power units Power generation at any given moment; For public power grid The amount of electricity purchased at any given time; For the park's integrated energy system The basic electrical load demand power at any given time; For electric refrigeration machine Input power at any given time; For electric energy storage system The charging power at any given time; For electric energy storage system Discharge power at any given moment; For combined heat and power units Constant heating; Gas-fired boiler Constant heating; For thermal energy storage systems The heat is released constantly; For the park's integrated energy system The heat load demand at any given time; For thermal energy storage systems Constant heat generation; Lithium bromide absorption chiller The amount of driving heat input at any given moment; For electric refrigeration Cooling output at any given time; Lithium bromide absorption chiller Cooling output at any given time; For cold energy storage system The amount of cooling released at any given moment; For the park's integrated energy system The cooling load requirement at any given time; For cold energy storage system The amount of cold storage at any given moment; This is the minimum allowable input power for the electric chiller; For electric refrigeration machine Input power at any given time; This refers to the maximum allowable input power of the electric chiller; This is the minimum allowable refrigeration output for a lithium bromide absorption chiller; Absorption chiller Cooling output at all times; For lithium bromide absorption chillers at all times The maximum allowable cooling output; This represents the maximum power output of the photovoltaic power generation unit. This represents the maximum power output of the wind power generation unit. This is the maximum power generation capacity of the combined heat and power unit; For the thermal efficiency of combined heat and power units; This is the maximum heat output of the gas-fired boiler; Thermal efficiency of gas-fired boilers; Gas-fired boiler Input power at any given time; For electric refrigeration machine The cooling energy efficiency ratio at any given time; The coefficient of performance (COP) of a lithium bromide absorption chiller; The refrigeration output jump variable of the lithium bromide absorption chiller; Lithium bromide absorption chiller Cooling output at time -1; For electric energy storage system Energy state at time +1; For electric energy storage system Energy state at any given moment; The charging efficiency of the energy storage system; For time step; The energy release efficiency of the energy storage system; For thermal energy storage systems Energy state at time +1; For thermal energy storage systems Energy state at any given moment; The charging efficiency of the thermal energy storage system; The energy release efficiency of the thermal energy storage system; For cold energy storage system Energy state at time +1; For cold energy storage system Energy state at any given moment; To improve the charging efficiency of the cold energy storage system; The energy release efficiency of the cold energy storage system; This represents the energy state limit of an electric energy storage system; This represents the upper limit of the energy state of an electric energy storage system. This is the upper limit of the charging power of an energy storage system. This is the upper limit of the discharge power of the energy storage system; This represents the energy state limit of a thermal energy storage system. This represents the upper limit of the energy state of a thermal energy storage system. This is the upper limit of the heat capacity of the thermal energy storage system; This is the upper limit of the heat release of the thermal energy storage system; This represents the energy state limit for cold energy storage systems. This represents the upper limit of the energy state of a cold energy storage system. This is the upper limit of the cooling capacity of the cold energy storage system; This is the upper limit of the charging and discharging capacity of the cold energy storage system.

[0025] The beneficial effects of the above-mentioned further solutions are: by establishing complete energy balance constraints, equipment operation constraints, and energy storage dynamic constraints, the physical feasibility and safety of the scheduling scheme are ensured.

[0026] Furthermore, the expression for the comfort deviation index of the user-side air conditioning cluster is as follows:

[0027]

[0028] in, For user-side air conditioning clusters Comfort deviation index at any given time; For time; The number of regions participating in the user-side air conditioning cluster scheduling; For area number indexing; For the first each region The actual temperature at that moment; For the first each region The set temperature at any given time; It is the absolute value; The lower limit of the comfortable temperature; This is the upper limit of the comfortable temperature range.

[0029] The beneficial effects of the above-mentioned further solutions are as follows: by introducing a comfort deviation index, the user's thermal comfort needs are quantified into an optimizable objective function, thus achieving a coordinated balance between energy efficiency and comfort.

[0030] Furthermore, the expression for the day-ahead scheduling optimization objective is:

[0031] in, The current scheduling optimization objective is to optimize the objective function value. Minimum; This is the electricity purchase cost coefficient; For public power grid The amount of electricity purchased at any given time; This is the fuel cost coefficient; for Fuel consumption at any given time; The operating and maintenance cost coefficient for equipment output; for Equipment output at all times; The unit carbon emission cost coefficient; for Equivalent carbon emissions at any given moment; The weighting coefficient for the comfort deviation index; for Comfort deviation index at any given time; This represents the total number of time steps within the scheduling period.

[0032] The beneficial effects of the above-mentioned further solutions are: by constructing a multi-objective optimization function that considers economic cost, carbon emissions, and comfort, a comprehensive balance between economy, low carbon emissions, and user comfort is achieved.

[0033] Furthermore, the upper-level optimization of the aforementioned day-ahead double-layer nested optimization model aims at source-load stability, minimizing the fluctuation of the power purchased by the public grid during the scheduling period; the upper-level optimization objective is to minimize the objective function value. The lower layer of the double-nested optimization model currently uses a comprehensive optimization objective. With minimization as the objective, non-cooperative game theory is used to coordinate the various air conditioning sub-clusters obtained from the user-side air conditioning cluster division, thereby achieving coordinated adjustment of response allocation and comfort deviation constraints among the sub-clusters:

[0034]

[0035]

[0036]

[0037]

[0038] in, This is the upper-level optimization objective of the current double-nested optimization model; This represents the total number of time steps within the scheduling period. For public power grid The amount of electricity purchased at any given time; This represents the average power purchased within the dispatching cycle; It is the absolute value; This is the lower-level comprehensive optimization objective of the current double-nested optimization model; for Weighting coefficients; This is the target economic cost value calculated from the cost of electricity and the cost of gas. for Weighting coefficients; This represents the equivalent carbon emissions generated from electricity purchases and gas consumption during the dispatch cycle. for Weighting coefficients; For the reason The target value of comfort deviation is obtained by accumulating the comfort deviation indices at different times; This is the electricity purchase cost coefficient; This is the fuel cost coefficient; for Fuel consumption at any given time; The operating and maintenance cost coefficient for equipment output; for Equipment output at all times; The unit carbon emission cost coefficient; for Equivalent carbon emissions at any given moment; for Comfort deviation index at any given time.

[0039] The beneficial effects of the above-mentioned further scheme are as follows: through the day-ahead double-layer nested optimization structure, the upper layer ensures the stability of the system source load, and the lower layer realizes the autonomous optimization of the air conditioning sub-cluster through game theory, thereby improving the global optimality of scheduling and the distributed decision-making capability.

[0040] Furthermore, the dynamic correction using an intraday rolling optimization and game-theoretic coordination model specifically involves: The system monitors the actual load and actual output of renewable energy in the park's integrated energy system in real time. When the deviation between the actual load or renewable energy output and the next day's baseline dispatch plan exceeds a preset threshold, or when the SOC of the energy storage unit exceeds a preset boundary threshold, the system triggers the intraday rolling optimization and game coordination model to dynamically correct the equipment output, purchased power, and energy storage charging and discharging power for the remaining time steps of the day.

[0041] The beneficial effects of the above-mentioned further solutions are: through the intraday rolling optimization mechanism, it is possible to respond in real time to fluctuations in renewable energy output and load changes, thereby improving the system's real-time adjustment capability and operational robustness.

[0042] Furthermore, the upper-level cooperative game participants in the intraday rolling optimization and game coordination model include grid interaction, energy storage, and total air conditioning power; the common optimization objective of the upper-level cooperative game participants is:

[0043] in, The goal of the upper-level comprehensive optimization is to minimize the target value. ; To step from the current time Until the end of the scheduling cycle The economic cost target value; To step from the current time Until the end of the scheduling cycle The target value for carbon emission costs; The lower layer of the intraday rolling optimization and game-theoretic coordination model independently optimizes each air conditioning sub-cluster under power constraints through non-cooperative game theory; the optimization objective of the lower layer of the intraday rolling optimization and game-theoretic coordination model is:

[0044] in, For carbon emission factor items; For the first Equivalent carbon emissions of each air conditioning sub-cluster in the current period; This is the coefficient for the comfort deviation term; For the first The comfort deviation values ​​of each air conditioning sub-cluster in the current time period; The weighting factor for the comfort deviation penalty of the air conditioning sub-cluster; For the first The constraint penalty factor for each air conditioning sub-cluster in the current iteration.

[0045] The beneficial effects of the above-mentioned further solutions are: through the intraday rolling optimization mechanism, it is possible to respond in real time to fluctuations in renewable energy output and load changes, thereby improving the system's real-time adjustment capability and operational robustness.

[0046] Furthermore, the improved leech swarm optimization algorithm specifically introduces an adaptive constraint penalty factor mechanism, an adaptive step size adjustment mechanism, a proportional self-adjustment mechanism, and an elite solution retention mechanism into the leech swarm optimization algorithm; the elite solution retention mechanism specifically involves prioritizing the retention of high-fitness individuals based on a set retention ratio when the target fluctuates or the individual feasibility decreases.

[0047] The beneficial effects of the above-mentioned further scheme are: by introducing mechanisms such as adaptive constraint penalty factor, self-adjustment of step size and preservation of elite solutions, the convergence speed and solution accuracy of the algorithm in high-dimensional non-convex problems are improved.

[0048] Furthermore, the expressions for the constraint penalty factor adaptive mechanism, the adaptive step size adjustment mechanism, and the proportional self-adjustment mechanism are as follows:

[0049]

[0050]

[0051] in, For the first +1 iteration constraint adjustment coefficient; For the first The constraint adjustment coefficient for the next iteration; It is a nonlinear adjustment function; For the first Constrain the default severity index in the next iteration; For the first The next iteration is used to control the step size adjustment of the search update magnitude; For the first +1 iterations are used to control the step size adjustment of the search update magnitude; This is a power correction boundary clipping function; To adjust the lower limit of the step size for the minimum amplitude limit; To adjust the upper limit of the step size for the maximum amplitude limit; To adjust the parameters; It is a natural constant; This represents the current group standard deviation. For the first The preset variance threshold in the next iteration; No. The step size decay coefficient of the next iteration; The threshold for determining convergence; This is the proportional adjustment coefficient.

[0052] The beneficial effects of the above-mentioned further scheme are: through a refined adaptive adjustment mechanism, the algorithm can dynamically adjust parameters according to the search state, avoid premature convergence, and improve the global search capability. Attached Figure Description

[0053] Figure 1 This is a flowchart of the method of the present invention.

[0054] Figure 2 This is a schematic diagram of the integrated energy system structure of the park.

[0055] Figure 3 This is a schematic diagram of an incentive-driven, collaborative optimization, and intelligent execution framework for the integrated energy system of a park.

[0056] Figure 4 This is a schematic diagram of a two-layer collaborative optimization framework for the integrated energy system of the park based on hybrid game theory.

[0057] Figure 5 This is a schematic diagram of the improved BLSO flexible load collaborative optimization solution module. Detailed Implementation

[0058] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0059] like Figure 1 and Figure 3 As shown, in one embodiment of the present invention, a method for scheduling a comprehensive energy system in a park that considers the energy consumption of a central air conditioning cluster includes: Establish a comprehensive energy system for the park, including power supply, cold and heat source equipment, energy storage units, and user-side air conditioning clusters; Construct a comfort deviation index for user-side air conditioning clusters; By selecting different historical load data, historical renewable energy output data, and historical meteorological data from the park's integrated energy system, and calling the trained deep learning-based load and renewable energy output prediction model, several time-series prediction samples for future scheduling cycles are generated; the time-series prediction samples include time-series data of electricity load, cooling load, and renewable energy output. Cluster analysis was performed on each time-series prediction sample to obtain several scene clusters; the time-series prediction sample closest to the center of each scene cluster was selected as the typical daily scene data of the corresponding scene cluster. Based on data from typical daily scenarios and comfort deviation indicators, daily scheduling optimization objectives corresponding to each typical daily scenario are constructed. Based on the corresponding daily scheduling optimization objectives, the scheduling results for each typical daily scenario are obtained by solving the problem through a daily double-nested optimization model. The load and renewable energy output prediction model trained based on the latest historical data is used to generate the next day's prediction time series. The distance between the next day's prediction time series and the time series prediction samples corresponding to each typical day scenario is measured, and the typical day scenario with the smallest distance is selected as the target scenario. The scheduling result corresponding to the target scenario is used as the baseline scheduling plan for the next day. During the operation of the park's integrated energy system the following day, the baseline scheduling plan for the following day is used as the initial plan, and dynamic correction is carried out using an intraday rolling optimization and game coordination model. The aforementioned day-ahead double-nested optimization model and intraday rolling optimization and game coordination model both use the improved leech swarm algorithm as the solver.

[0060] In this embodiment, the trained prediction model is invoked to generate predicted time series for future scheduling cycles based on different historical inputs, and time series prediction samples are constructed. Based on the latest historical data collected before the scheduling solution time, the aforementioned load and renewable energy output prediction model is invoked to obtain the load and renewable energy output prediction sequences for the next day's scheduling cycle.

[0061] In this embodiment, the purpose of this invention is to overcome the problems of insufficient time-scale coordination, limited flexibility of central air conditioning, and low operational robustness in existing integrated energy systems for industrial parks. It provides a scheduling method for integrated energy systems of central air conditioning clusters based on a two-layer nested collaborative optimization approach. Through a day-ahead to intraday dual-time-scale collaborative structure, the overall economic efficiency and local flexibility of the integrated energy system for industrial parks can be unified. Based on a central air conditioning load model and a game-theoretic coordination mechanism, the air conditioning group is transformed from a passive response unit into an active regulation subject, effectively increasing the regulation performance of the integrated energy system for industrial parks. Simultaneously, this invention combines an adaptive intelligent optimization algorithm to achieve rapid solution and stable convergence under multi-objective and multi-constraint conditions, ensuring that the scheduling scheme possesses high economic efficiency, high feasibility, and high robustness under conditions of high proportion of renewable energy.

[0062] In this embodiment, a prediction model based on deep learning is used to generate future load and renewable energy output scenarios (using methods such as Transformer and TimeGAN), and a clustering algorithm (such as K-means) is used to extract representative typical days, providing multi-scenario references for day-ahead optimization.

[0063] In this embodiment, a two-layer nested optimization structure is adopted in the day-ahead phase: the upper layer aims at energy stability, coordinating the power distribution and fluctuation suppression of the three energy flows (electricity, heat, and cooling); the lower layer comprehensively minimizes operating costs, carbon emissions, and comfort deviations, achieving multi-objective coordination. Through interactive solutions between the upper and lower layers, the day-ahead optimal solution that balances global economics and energy efficiency is obtained. In the intraday phase, rolling re-optimization is performed based on real-time load, renewable energy output, and energy storage status deviations. The upper layer uses cooperative game theory to coordinate the total system cost and user comfort, while the lower layer optimizes the energy efficiency of each air conditioning sub-cluster using non-cooperative game theory, realizing a hybrid game mechanism of "upper-layer coordination—lower-layer response," thereby achieving multi-level collaboration. Under multi-objective and multi-constraint conditions, an improved Bloodsucker Leech Swarm Optimization (BLSO) algorithm is used to uniformly encode and solve continuous and discrete decision variables. This algorithm improves convergence speed and solution feasibility through adaptive search step size and elite solution retention strategies. The system ultimately outputs 24-hour operation plans for each device, energy storage charging and discharging strategies, and air conditioning group control commands, enabling dynamic execution and rolling adjustments.

[0064] In this embodiment, the present invention employs a day-ahead and intraday double-layer nested mechanism, enabling the park's integrated energy system to achieve optimal economic efficiency over a long timescale and possess real-time response and fluctuation compensation capabilities over a short timescale, effectively improving the coordination and stability of system scheduling. By using dual constraints of comfort temperature zones and power settings, the central air conditioning system is transformed from a rigid load into an adjustable flexible unit, helping the park's integrated energy system achieve peak shaving and valley filling and demand response, balancing comfort and energy efficiency. By comprehensively considering the interaction and conversion between electrical, thermal, and cooling energy flows, and combining rolling optimization and scenario correction mechanisms, the impact of renewable energy output fluctuations in the park's integrated energy system on system scheduling performance is effectively reduced, improving the system's energy stability.

[0065] The power supply side is used for power supply and includes photovoltaic power generation units, wind power generation units, gas boilers, combined heat and power units and the public power grid; The aforementioned heat and cold source equipment is used to realize the mutual conversion between electrical energy, thermal energy and cold energy, including electric chillers and lithium bromide absorption chillers; The energy storage unit is used to realize energy time migration and peak shaving and valley filling, including electrical energy storage system, thermal energy storage system and cold energy storage system.

[0066] In this embodiment, a comprehensive energy system model for the park is established, including power supply side, cold and heat source equipment, energy storage units, and user-side air conditioning clusters. The system includes photovoltaic power generation units, gas boilers, combined heat and power units, lithium bromide absorption chillers, electric chillers, electric / heat / cold energy storage equipment, and multiple central air conditioning sub-clusters. To achieve flexible modeling, the comfort temperature range, cooling power limit, start-stop constraints, and energy efficiency ratio (EER) curves of each sub-cluster are defined.

[0067] In this embodiment, as Figure 2 As shown, the system is structured as follows: Power supply side: including photovoltaic (PV) power generation units, wind power generation units (WT), gas boilers (GB), combined heat and power (CHP) units, and the public grid (Grid); Energy conversion equipment: including electric chillers (EC), lithium bromide absorption chillers (LBAC), and electric heat pumps (HP), realizing the mutual conversion between electrical energy, thermal energy, and cold energy; Energy storage unit: including electric energy storage system (EES), thermal energy storage system (TES), and cold energy storage system (CES), used to realize energy time migration and peak shaving and valley filling; The load side includes three categories: electrical load, heat load, and cooling load. Electrical load mainly consists of rigid electrical consumption such as lighting, office equipment, and power equipment; heat load includes heating and domestic hot water systems; and cooling load mainly consists of multiple central air conditioning clusters and chiller plant loads. By dividing the system by building or region, these subsystems form a comprehensive energy network that integrates multi-energy complementarity and energy interaction.

[0068] The operational constraints of the park's integrated energy system include energy balance constraints, energy equipment operational constraints, and energy storage unit dynamic constraints.

[0069]

[0070]

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] in, Photovoltaic power generation unit Power generation at any given moment; Wind power generation unit Power generation at any given moment; For combined heat and power units Power generation at any given moment; For public power grid The amount of electricity purchased at any given time; For the park's integrated energy system The basic electrical load demand power at any given time; For electric refrigeration machine Input power at any given time; For electric energy storage system The charging power at any given moment; For electric energy storage system Discharge power at any given moment; For combined heat and power units Constant heating; Gas-fired boiler Constant heating; For thermal energy storage systems The heat release at all times; For the park's integrated energy system The heat load demand at any given time; For thermal energy storage systems Constant heat generation; Lithium bromide absorption chiller The amount of driving heat input at any given moment; For electric refrigeration machine Cooling output at any given time; Lithium bromide absorption chiller Cooling output at any given time; For cold energy storage system The amount of cooling released at any given moment; For the park's integrated energy system The cooling load requirement at any given time; For cold energy storage system The amount of cold storage at any given moment; This is the minimum allowable input power for the electric chiller; For electric refrigeration machine Input power at any given time; This refers to the maximum allowable input power of the electric chiller; This is the minimum allowable refrigeration output for a lithium bromide absorption chiller; Absorption chiller Cooling output at all times; For lithium bromide absorption chillers at all times The maximum allowable cooling output; This represents the maximum power output of the photovoltaic power generation unit. This represents the maximum power output of the wind power generation unit. This is the maximum power generation capacity of the combined heat and power unit; For the thermal efficiency of combined heat and power units; This is the maximum heat output of the gas-fired boiler; Thermal efficiency of gas-fired boilers; Gas-fired boiler Input power at any given time; For electric refrigeration machine The cooling energy efficiency ratio at any given time; The coefficient of performance (COP) of a lithium bromide absorption chiller; The refrigeration output jump variable of the lithium bromide absorption chiller; Lithium bromide absorption chiller Cooling output at time -1; For electric energy storage system Energy state at time +1; For electric energy storage system Energy state at any given moment; The charging efficiency of the energy storage system; For time step; The energy release efficiency of the energy storage system; For thermal energy storage systems Energy state at time +1; For thermal energy storage systems Energy state at any given moment; The charging efficiency of the thermal energy storage system; The energy release efficiency of the thermal energy storage system; For cold energy storage system Energy state at time +1; For cold energy storage system Energy state at any given moment; To improve the charging efficiency of the cold energy storage system; The energy release efficiency of the cold energy storage system; This represents the energy state limit for an electric energy storage system. This represents the upper limit of the energy state of an electric energy storage system. This is the upper limit of the charging power of an energy storage system. This is the upper limit of the discharge power of the energy storage system; This represents the energy state limit of a thermal energy storage system. This represents the upper limit of the energy state of a thermal energy storage system. This is the upper limit of the heat capacity of the thermal energy storage system; This is the upper limit of the heat release of the thermal energy storage system; This represents the energy state limit for cold energy storage systems. This represents the upper limit of the energy state of a cold energy storage system. This is the upper limit of the cooling capacity of the cold energy storage system; This is the upper limit of the charging and discharging capacity of the cold energy storage system.

[0080] The expression for the comfort deviation index of the user-side air conditioning cluster is:

[0081]

[0082] in, For user-side air conditioning clusters Comfort deviation index at any given time; For time; The number of regions participating in the user-side air conditioning cluster scheduling; For area number indexing; For the first each region The actual temperature at that moment; For the first each region The set temperature at any given time; It is the absolute value; The lower limit of the comfortable temperature; This is the upper limit of the comfortable temperature range.

[0083] The expression for the day-ahead scheduling optimization objective is:

[0084] in, The current scheduling optimization objective is to optimize the objective function value. Minimum; This is the electricity purchase cost coefficient; For public power grid The amount of electricity purchased at any given time; This is the fuel cost coefficient; for Fuel consumption at any given time; The operating and maintenance cost coefficient for equipment output; for Equipment output at all times; The unit carbon emission cost coefficient; for Equivalent carbon emissions at any given moment; The weighting coefficient for the comfort deviation index; for Comfort deviation index at any given time; This represents the total number of time steps within the scheduling period.

[0085] The upper-level optimization of the aforementioned two-layer nested optimization model aims at source-load stability and minimizes the fluctuation of the power purchased by the public grid during the scheduling period; the upper-level optimization objective is to minimize the objective function value. The lower layer of the double-nested optimization model currently uses a comprehensive optimization objective. With minimization as the objective, non-cooperative game theory is used to coordinate the various air conditioning sub-clusters obtained from the user-side air conditioning cluster division, thereby achieving coordinated adjustment of response allocation and comfort deviation constraints among the sub-clusters:

[0086]

[0087]

[0088]

[0089]

[0090] in, This is the upper-level optimization objective of the current double-nested optimization model; This represents the total number of time steps within the scheduling period. For public power grid The amount of electricity purchased at any given time; This represents the average power purchased within the dispatching cycle; It is the absolute value; This is the lower-level comprehensive optimization objective of the current double-nested optimization model; for Weighting coefficients; This is the target economic cost value calculated from the cost of electricity and the cost of gas. for Weighting coefficients; This represents the equivalent carbon emissions generated from electricity purchases and gas consumption during the dispatch cycle. for Weighting coefficients; For the reason The target value of comfort deviation is obtained by accumulating the comfort deviation indices at different times; This is the electricity purchase cost coefficient; This is the fuel cost coefficient; for Fuel consumption at any given time; The operating and maintenance cost coefficient for equipment output; for Equipment output at all times; The unit carbon emission cost coefficient; for Equivalent carbon emissions at any given moment; for Comfort deviation index at any given time.

[0091] The method of using intraday rolling optimization and game-theoretic coordination model for dynamic correction is as follows: The system monitors the actual load and actual output of renewable energy in the park's integrated energy system in real time. When the deviation between the actual load or renewable energy output and the next day's baseline dispatch plan exceeds a preset threshold, or when the SOC of the energy storage unit exceeds a preset boundary threshold, the system triggers the intraday rolling optimization and game coordination model to dynamically correct the equipment output, purchased power, and energy storage charging and discharging power for the remaining time steps of the day.

[0092] The upper-level cooperative game participants in the intraday rolling optimization and game coordination model include grid interaction, energy storage, and total air conditioning power; the common optimization objective of the upper-level cooperative game participants is:

[0093] in, The goal of the upper-level comprehensive optimization is to minimize the target value. ; To step from the current time Until the end of the scheduling cycle The economic cost target value; To step from the current time Until the end of the scheduling cycle The target value for carbon emission costs; The lower layer of the intraday rolling optimization and game-theoretic coordination model independently optimizes each air conditioning sub-cluster under power constraints through non-cooperative game theory; the optimization objective of the lower layer of the intraday rolling optimization and game-theoretic coordination model is:

[0094] in, For carbon emission factor items; For the first Equivalent carbon emissions of each air conditioning sub-cluster in the current period; This is the coefficient for the comfort deviation term; For the first The comfort deviation values ​​of each air conditioning sub-cluster in the current time period; The weighting factor for the comfort deviation penalty of the air conditioning sub-cluster; For the first The constraint penalty factor for each air conditioning sub-cluster in the current iteration.

[0095] In this embodiment, as Figure 4 As shown, the interaction mechanism involves the upper layer allocating power → the lower layer providing feedback → iterative convergence between the upper and lower layers → forming an equilibrium solution → distributing it to each sub-cluster for execution. Through a hybrid game of "cooperative + non-cooperative" approaches, the overall system efficiency is ensured while maintaining the autonomy of sub-clusters and user comfort.

[0096] The improved leech swarm optimization algorithm specifically introduces an adaptive constraint penalty factor mechanism, an adaptive step size adjustment mechanism, a proportional self-adjustment mechanism, and an elite solution retention mechanism into the leech swarm optimization algorithm; the elite solution retention mechanism specifically involves prioritizing the retention of high-fitness individuals based on a set retention ratio when the target fluctuates or the individual feasibility decreases.

[0097] The expressions for the constraint penalty factor adaptive mechanism, the adaptive step size adjustment mechanism, and the proportional self-adjustment mechanism are as follows:

[0098]

[0099]

[0100] in, For the first +1 iteration constraint adjustment coefficient; For the first The constraint adjustment coefficient for the next iteration; It is a nonlinear adjustment function; For the first Constrain the default severity index in the next iteration; For the first The next iteration is used to control the step size adjustment of the search update magnitude; For the first +1 iterations are used to control the step size adjustment of the search update magnitude; This is a power correction boundary clipping function; To adjust the lower limit of the step size for the minimum amplitude limit; To adjust the upper limit of the step size for the maximum amplitude limit; To adjust the parameters; It is a natural constant; This represents the current group standard deviation. For the first The preset variance threshold in the next iteration; No. The step size decay coefficient of the next iteration; The threshold for determining convergence; This is the proportional adjustment coefficient.

[0101] In this embodiment, as Figure 5 As shown, to address the complex coupled problem of multiple objectives and constraints, this invention introduces an adaptive improvement mechanism based on the traditional BLSO algorithm. The improved algorithm, through dynamic parameter adjustment and swarm cooperative search, enhances the convergence speed and solution set stability of multi-energy system optimization, specifically including the following four aspects: ① Constraint penalty factor adaptive mechanism: In response to the problem of fixed penalty terms and low constraint feasibility in traditional algorithms, this invention dynamically adjusts the penalty factor based on the degree of constraint violation; when the algorithm deviates from the constraint boundary during iteration, the penalty weight is automatically increased to encourage the individual to return to the feasible region; as the solution gradually approaches the feasible space, the penalty coefficient decreases gradually, thereby maintaining stable convergence.

[0102] ② Adaptive step size adjustment mechanism: To balance global exploration and local convergence capabilities, a dynamic step size adjustment strategy based on search success rate is introduced; the algorithm maintains a large step size in the early stage to enhance global search, and automatically reduces the step size in the later stage to improve local convergence accuracy, thereby achieving a "fast first, stable later" convergence process.

[0103] ③ Global / Local Proportion Self-Adjustment Mechanism: To prevent the algorithm from getting stuck in local optima, a proportion adaptive mechanism based on population variance is introduced to dynamically balance global search and local exploration; the algorithm automatically increases the proportion of global exploration; when the population difference increases, the algorithm strengthens the local search to ensure the diversity and convergence balance of the search process.

[0104] ④ Elite Solution Retention Mechanism: When the objective fluctuates or the feasibility of an individual decreases, the algorithm adopts an elite retention strategy, prioritizing the retention of high-fitness individuals to prevent the optimal solution from being lost during iteration. By controlling the retention ratio, the stability of high-quality solutions is ensured even under changes in the weights of multiple objectives. This mechanism enables the optimization framework to achieve a dynamic logic of "feasibility first, then optimization, and finally stable convergence," making it suitable for scheduling optimization under multiple scenarios and constraints.

Claims

1. A method for scheduling a comprehensive energy system in a park that considers the energy consumption of a central air conditioning cluster, characterized in that, include: Establish a comprehensive energy system for the park, including power supply, cold and heat source equipment, energy storage units, and user-side air conditioning clusters; Construct a comfort deviation index for user-side air conditioning clusters; By selecting different historical load data, historical renewable energy output data, and historical meteorological data from the park's integrated energy system, and calling the trained deep learning-based load and renewable energy output prediction model, several time-series prediction samples for future scheduling cycles are generated; the time-series prediction samples include time-series data of electricity load, cooling load, and renewable energy output. Cluster analysis was performed on each time-series prediction sample to obtain several scene clusters; the time-series prediction sample closest to the center of each scene cluster was selected as the typical daily scene data of the corresponding scene cluster. Based on data from typical daily scenarios and comfort deviation indicators, daily scheduling optimization objectives corresponding to each typical daily scenario are constructed. Based on the corresponding daily scheduling optimization objectives, the scheduling results for each typical daily scenario are obtained by solving the problem through a daily double-nested optimization model. The load and renewable energy output prediction model trained based on the latest historical data is used to generate the next day's prediction time series. The distance between the next day's prediction time series and the time series prediction samples corresponding to each typical day scenario is measured, and the typical day scenario with the smallest distance is selected as the target scenario. The scheduling result corresponding to the target scenario is used as the baseline scheduling plan for the next day. During the operation of the park's integrated energy system the following day, the baseline scheduling plan for the following day is used as the initial plan, and dynamic correction is carried out using an intraday rolling optimization and game coordination model. The aforementioned day-ahead double-nested optimization model and intraday rolling optimization and game coordination model both use the improved leech swarm algorithm as the solver.

2. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The power supply side is used for power supply and includes photovoltaic power generation units, wind power generation units, gas boilers, combined heat and power units and the public power grid; The aforementioned heat and cold source equipment is used to realize the mutual conversion between electrical energy, thermal energy and cold energy, including electric chillers and lithium bromide absorption chillers; The energy storage unit is used to realize energy time migration and peak shaving and valley filling, including electrical energy storage system, thermal energy storage system and cold energy storage system.

3. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The operational constraints of the park's integrated energy system include energy balance constraints, energy equipment operational constraints, and energy storage unit dynamic constraints. in, Photovoltaic power generation unit Power generation at any given moment; Wind power generation unit Power generation at any given moment; For combined heat and power units Power generation at any given moment; For public power grid The amount of electricity purchased at any given time; For the park's integrated energy system The basic electrical load demand power at any given time; For electric refrigeration Input power at any given time; For electric energy storage system The charging power at any given time; For electric energy storage system Discharge power at any given moment; For combined heat and power units Constant heating; Gas-fired boiler Constant heating; For thermal energy storage systems The heat is released constantly; For the park's integrated energy system The heat load demand at any given time; For thermal energy storage systems Constant heat generation; Lithium bromide absorption chiller The amount of driving heat input at any given moment; For electric refrigeration Cooling output at any given time; Lithium bromide absorption chiller Cooling output at any given time; For cold energy storage system The amount of cooling released at any given moment; For the park's integrated energy system The cooling load requirement at any given time; For cold energy storage system The amount of cold storage at any given moment; This is the minimum allowable input power for the electric chiller; For electric refrigeration Input power at any given time; This refers to the maximum allowable input power of the electric chiller; This is the minimum allowable refrigeration output for a lithium bromide absorption chiller; Absorption chiller Cooling output at all times; For lithium bromide absorption chillers at all times The maximum allowable cooling output; This represents the maximum power output of the photovoltaic power generation unit. This represents the maximum power output of the wind power generation unit. This is the maximum power generation capacity of the combined heat and power unit; For the thermal efficiency of combined heat and power units; This is the maximum heat output of the gas-fired boiler; Thermal efficiency of gas-fired boilers; Gas-fired boiler Input power at any given time; For electric refrigeration The cooling energy efficiency ratio at any given time; The coefficient of performance (COP) of a lithium bromide absorption chiller; The refrigeration output jump variable of the lithium bromide absorption chiller; Lithium bromide absorption chiller Cooling output at time -1; For electric energy storage system Energy state at time +1; For electric energy storage system Energy state at any given moment; The charging efficiency of the energy storage system; For time step; The energy release efficiency of the energy storage system; For thermal energy storage systems Energy state at time +1; For thermal energy storage systems Energy state at any given moment; The charging efficiency of the thermal energy storage system; The energy release efficiency of the thermal energy storage system; For cold energy storage system Energy state at time +1; For cold energy storage system Energy state at any given moment; To improve the charging efficiency of the cold energy storage system; The energy release efficiency of the cold energy storage system; This represents the energy state limit for an electric energy storage system. This represents the upper limit of the energy state of an electric energy storage system. This is the upper limit of the charging power of an energy storage system. This is the upper limit of the discharge power of the energy storage system; This represents the energy state limit of a thermal energy storage system. This represents the upper limit of the energy state of a thermal energy storage system. This is the upper limit of the heat capacity of the thermal energy storage system; This is the upper limit of the heat release of the thermal energy storage system; This represents the energy state limit for cold energy storage systems. This represents the upper limit of the energy state of a cold energy storage system. This is the upper limit of the cooling capacity of the cold energy storage system; This is the upper limit of the charging and discharging capacity of the cold energy storage system.

4. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The expression for the comfort deviation index of the user-side air conditioning cluster is: in, For user-side air conditioning clusters Comfort deviation index at any given time; For time; The number of regions participating in the user-side air conditioning cluster scheduling; For area number indexing; For the first each region The actual temperature at that moment; For the first each region The set temperature at any given time; It is the absolute value; The lower limit of the comfortable temperature; This is the upper limit of the comfortable temperature range.

5. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The expression for the day-ahead scheduling optimization objective is: in, The current scheduling optimization objective is to optimize the objective function value. Minimum; This is the electricity purchase cost coefficient; For public power grid The amount of electricity purchased at any given time; This is the fuel cost coefficient; for Fuel consumption at any given time; The operating and maintenance cost coefficient for equipment output; for Equipment output at all times; The unit carbon emission cost coefficient; for Equivalent carbon emissions at any given moment; The weighting coefficient for the comfort deviation index; for Comfort deviation index at any given time; This represents the total number of time steps within the scheduling period.

6. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The upper-level optimization of the aforementioned two-layer nested optimization model aims at source-load stability and minimizes the fluctuation of the power purchased by the public grid during the scheduling period; the upper-level optimization objective is to minimize the objective function value. The lower layer of the double-nested optimization model currently uses a comprehensive optimization objective. With minimization as the objective, non-cooperative game theory is used to coordinate the various air conditioning sub-clusters obtained from the user-side air conditioning cluster division, thereby achieving coordinated adjustment of response allocation and comfort deviation constraints among the sub-clusters: in, This is the upper-level optimization objective of the current double-nested optimization model; This represents the total number of time steps within the scheduling period. For public power grid The amount of electricity purchased at any given time; This represents the average power purchased within the dispatching cycle; It is the absolute value; This is the lower-level comprehensive optimization objective of the current double-nested optimization model; for Weighting coefficients; This is the target economic cost value calculated from the cost of electricity and the cost of gas. for Weighting coefficients; This represents the equivalent carbon emissions generated from electricity purchases and gas consumption during the dispatch cycle. for Weighting coefficients; For the reason The target value of comfort deviation is obtained by accumulating the comfort deviation indices at different times; This is the electricity purchase cost coefficient; This is the fuel cost coefficient; for Fuel consumption at any given time; The operating and maintenance cost coefficient for equipment output; for Equipment output at all times; The unit carbon emission cost coefficient; for Equivalent carbon emissions at any given moment; for Comfort deviation index at any given time.

7. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The method of using intraday rolling optimization and game-theoretic coordination model for dynamic correction is as follows: The system monitors the actual load and actual output of renewable energy in the park's integrated energy system in real time. When the deviation between the actual load or renewable energy output and the next day's baseline dispatch plan exceeds a preset threshold, or when the SOC of the energy storage unit exceeds a preset boundary threshold, the system triggers the intraday rolling optimization and game coordination model to dynamically correct the equipment output, purchased power, and energy storage charging and discharging power for the remaining time steps of the day.

8. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 7, is characterized in that... The upper-level cooperative game participants in the intraday rolling optimization and game coordination model include grid interaction, energy storage, and total air conditioning power; the common optimization objective of the upper-level cooperative game participants is: in, The goal of the upper-level comprehensive optimization is to minimize the target value. ; To step from the current time Until the end of the scheduling cycle The economic cost target value; To step from the current time Until the end of the scheduling cycle The target value for carbon emission costs; The lower layer of the intraday rolling optimization and game-theoretic coordination model independently optimizes each air conditioning sub-cluster under power constraints through non-cooperative game theory; the optimization objective of the lower layer of the intraday rolling optimization and game-theoretic coordination model is: in, For carbon emission factor items; For the first Equivalent carbon emissions of each air conditioning sub-cluster in the current period; This is the coefficient for the comfort deviation term; For the first The comfort deviation values ​​of each air conditioning sub-cluster in the current time period; The weighting factor for the comfort deviation penalty of the air conditioning sub-cluster; For the first The constraint penalty factor for each air conditioning sub-cluster in the current iteration.

9. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 1, is characterized in that... The improved leech swarm optimization algorithm specifically introduces an adaptive constraint penalty factor mechanism, an adaptive step size adjustment mechanism, a proportional self-adjustment mechanism, and an elite solution retention mechanism into the leech swarm optimization algorithm; the elite solution retention mechanism specifically involves prioritizing the retention of high-fitness individuals based on a set retention ratio when the target fluctuates or the individual feasibility decreases.

10. The method for scheduling a comprehensive energy system in a park considering the energy consumption of a central air conditioning cluster, as described in claim 9, is characterized in that... The expressions for the constraint penalty factor adaptive mechanism, the adaptive step size adjustment mechanism, and the proportional self-adjustment mechanism are as follows: in, For the first +1 iteration constraint adjustment coefficient; For the first The constraint adjustment coefficient for the next iteration; It is a nonlinear adjustment function; For the first Constrain the default severity index in the next iteration; For the first The next iteration is used to control the step size adjustment of the search update magnitude; For the first +1 iterations are used to control the step size adjustment of the search update magnitude; This is a power correction boundary clipping function; To adjust the lower limit of the step size for the minimum amplitude limit; To adjust the upper limit of the step size for the maximum amplitude limit; To adjust the parameters; It is a natural constant; This represents the current group standard deviation. For the first The preset variance threshold in the next iteration; No. The step size decay coefficient of each iteration; The threshold for determining convergence; This is the proportional adjustment coefficient.

Images