Park resource scheduling method and device, equipment and medium

Abstract

The invention relates to a park resource scheduling method and device, equipment and a medium, and relates to the technical field of resource scheduling. The method comprises the following steps: constructing a scheduling model corresponding to a park based on basic data corresponding to each resource in the park and a confidence gap decision theory; wherein the scheduling model performs grade classification on the resources according to the response condition of each resource in the park, and determines an adjustment mode corresponding to each resource; and standardizing and solving the scheduling model to obtain a resource scheduling result corresponding to the park. By adopting the method, a resource scheduling result matched with the park can be obtained.

Description

Technical Field

[0001] This application relates to the field of resource scheduling technology, and in particular to a method, apparatus, equipment and medium for scheduling resources in a park. Background Technology

[0002] With the continuous development of new power systems, integrated energy systems based in industrial parks are gradually becoming key carriers for improving energy efficiency and promoting the local consumption of distributed clean energy. These parks integrate various energy elements such as photovoltaics, energy storage, ground source heat pumps, and flexible loads, exhibiting typical characteristics of multi-energy coupling between source, grid, load, and storage. However, the inherent intermittency and strong volatility of renewable energy sources such as photovoltaics pose serious challenges to the real-time energy balance and stable operation within these parks.

[0003] Currently, there is considerable research and practice regarding the optimized operation of integrated energy systems in industrial parks. Most existing technical solutions revolve around deterministic scenarios, that is, economic dispatching or optimal allocation under the premise that photovoltaic output and load demand are known or assumed to be fixed values. However, these solutions still have significant limitations, ultimately leading to a disconnect between the model and the actual system, and the resulting resource dispatching results not matching the actual conditions of the industrial park.

[0004] Therefore, there is an urgent need for a resource allocation method that takes into account the specific circumstances of the park. Summary of the Invention

[0005] This application provides a method, apparatus, equipment, and medium for resource scheduling in a park, which can obtain resource scheduling results that match the park.

[0006] To achieve the above objectives, this application adopts the following technical solution: Firstly, this application provides a method for scheduling park resources, including: Based on the basic data of each resource in the park and the confidence gap decision theory, a scheduling model for the park is constructed. The scheduling model classifies resources according to their response status and determines the adjustment mode for each resource. The scheduling model is standardized and solved to obtain the resource scheduling results for the park.

[0007] In one embodiment, based on the basic data corresponding to each resource in the park and the confidence gap decision theory, a scheduling model corresponding to the park is constructed, including: Based on the basic data of each resource in the park and the confidence gap decision theory, the objective function and physical constraints of the scheduling model are constructed. Based on the objective function, physical constraints, and adjustment variables, a scheduling model corresponding to the park is constructed.

[0008] In one embodiment, based on the basic data corresponding to each resource in the park and the confidence gap decision theory, the objective function and physical constraints of the scheduling model are constructed, including: Based on resource operation parameters and the response demand of the electricity market at different times, the regulation level corresponding to each resource is determined; Based on confidence gap decision theory and the adjustment levels corresponding to each resource, an objective function for the scheduling model is constructed. Based on the basic data, the supply and demand balance relationship and operating characteristics of each resource are determined, and the physical constraints of the scheduling model are constructed. Based on the objective function, physical constraints, and adjustment variables, a scheduling model corresponding to the park is constructed.

[0009] In one embodiment, it includes: Based on photovoltaic power output forecast data, load data, resource operation parameters, and preset photovoltaic forecast error confidence levels, the confidence fluctuation range of photovoltaic power output is determined. A bias compensation mechanism is constructed based on the confidence fluctuation range and the adjustment variable.

[0010] In one embodiment, the regulation level corresponding to each resource is determined based on resource operating parameters and the response demand of the electricity market at different times, including: Based on the resource operation parameters, determine the response time and adjustment range for each resource; Based on the response demand of the electricity market at different times, and the response time and adjustment range of each resource, the adjustment level corresponding to each resource is determined.

[0011] In one embodiment, it includes: Based on the adjustment level, determine the corresponding participation period for each resource.

[0012] In one embodiment, the scheduling model is standardized and solved to obtain the resource scheduling results corresponding to the park, including: The scheduling model is linearized by nonlinear constraints, and the linearized scheduling model is solved to obtain the resource scheduling results corresponding to the park. The resource scheduling results include detailed resource hierarchical scheduling, photovoltaic uncertainty response results, change data of adjustment variables, and satisfaction of operational constraints.

[0013] Secondly, this application provides a park resource scheduling device, comprising: The model building module is used to construct a scheduling model for the park based on the basic data of each resource in the park and the confidence gap decision theory. The scheduling model classifies resources according to their response status and determines the adjustment mode for each resource. The scheduling module is used to standardize and solve the scheduling model to obtain the resource scheduling results for the park.

[0014] Thirdly, this application provides a computing device, including a memory and a processor; The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of the first aspects.

[0015] Fourthly, this application provides a computer-readable storage medium for storing a computer program for performing the method as described in any one of the first aspects.

[0016] Fifthly, this application provides a computer program product comprising one or more computer instructions, wherein when the computer instructions are executed by a computer, the computer performs the method as described in any one of the first aspects.

[0017] As can be seen from the above technical solution, this application has at least the following beneficial effects: In this application, based on the basic data of each resource in the park and the confidence gap decision theory, a scheduling model is constructed to classify resources according to their response status and determine the corresponding adjustment mode for each resource, thus determining the degree to which each resource in the park can participate in resource scheduling. Furthermore, the scheduling model is standardized and solved to obtain the resource scheduling results corresponding to the park, achieving perfect scheduling of park resources. This solution, by introducing a scheduling model that includes resource levels and adjustment modes, maximizes the value of resources throughout their entire lifecycle, obtains a scheduling strategy that precisely matches the market structure, and ultimately achieves resource scheduling results that match the park.

[0018] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0019] Figure 1 This is an application environment diagram of a park resource scheduling method provided in the embodiments of this application; Figure 2 This is a flowchart illustrating a park resource scheduling method provided in an embodiment of this application; Figure 3 This is a schematic diagram illustrating a process for constructing a scheduling model provided in an embodiment of this application; Figure 4 A structural block diagram of a park resource scheduling device is provided in the embodiments of this application; Figure 5 This is an internal structural diagram of a computer device provided in the embodiments of the application. Detailed Implementation

[0020] The terms "first," "second," and "third," etc., used in this application specification and accompanying drawings are used to distinguish different objects, not to limit a specific order.

[0021] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0022] With the continuous development of new power systems, integrated energy systems based in industrial parks are gradually becoming key carriers for improving energy efficiency and promoting the local consumption of distributed clean energy. These parks integrate various energy elements such as photovoltaics, energy storage, ground source heat pumps, and flexible loads, exhibiting typical characteristics of multi-energy coupling between source, grid, load, and storage. However, the inherent intermittency and strong volatility of renewable energy sources such as photovoltaics pose severe challenges to the real-time energy balance and stable operation within these parks. Furthermore, how to efficiently integrate the dispersed flexible resources within the system to create value in supporting grid stability and participating in the electricity market has become a core issue that urgently needs to be addressed in current industrial development and technological innovation.

[0023] Currently, there is considerable research and practice regarding the optimized operation of integrated energy systems in industrial parks. Most existing technical solutions revolve around deterministic scenarios, i.e., economic dispatching or optimal allocation under the premise that photovoltaic output and load demand are known or assumed to be fixed values. Some advanced research has begun to introduce uncertainties, such as using stochastic optimization or simple robust optimization methods to handle photovoltaic prediction errors, or making decisions in response to electricity price fluctuations. Furthermore, some solutions have also taken into account the differences in resource characteristics within the system, attempting to control rapidly adjustable resources such as energy storage separately. These methods provide a preliminary theoretical foundation and technical pathway for energy management in industrial parks.

[0024] Despite some progress in existing research, significant limitations remain. First, most methods handle uncertainty in a coarse manner, either relying on precise probability distributions (which are difficult to obtain) or employing overly conservative interval estimations, failing to adequately balance system economy and robustness. Second, existing schemes generally lack a resource scheduling mechanism finely coupled with the electricity market (especially the spot market divided into day-ahead, intraday, and real-time markets), failing to systematically classify, categorize, and coordinate the allocation of flexible resources (such as batteries, ice storage, and adjustable loads) with varying response characteristics within the industrial park according to market requirements for response speeds at different time scales. Finally, many optimization models neglect nonlinear constraints such as the mutual exclusivity of equipment operating states (e.g., heating / cooling cannot occur simultaneously), or are difficult to solve, leading to a disconnect between the model and the actual system and weak engineering feasibility. Therefore, developing a comprehensive energy system optimization method for industrial parks that can overcome these shortcomings and possess robustness, economy, and engineering practicality is of great significance.

[0025] To make the technical solution of this application clearer and easier to understand, the application scenarios of the technical solution of this application are described below with reference to the accompanying drawings. Figure 1 As shown in the figure, this figure is a schematic diagram of an application scenario provided by an embodiment of this application.

[0026] In this application scenario, server 104 calculates resource scheduling results based on basic data and confidence gap decision theory, and transmits the resource scheduling results to terminal 102 through the communication network so that terminal 102 can display the resource scheduling results.

[0027] To make the technical solution of this application clearer and easier to understand, the following describes a park resource scheduling method provided by an embodiment of this application, in conjunction with the above application scenarios. For example... Figure 2 As shown in the figure, this figure is a flowchart of a park resource scheduling method provided in an embodiment of this application.

[0028] S201. Based on the basic data of each resource in the park and the confidence gap decision theory, construct the scheduling model corresponding to the park.

[0029] The basic data includes resource types, the connection between resources and energy networks, photovoltaic output forecast data, load data, resource operating parameters, and grid purchase prices. Confidence gap decision theory is an improved uncertainty decision-making method that quantifies the risk of uncertain events by introducing probabilistic confidence constraints (such as setting a 95% confidence level). The scheduling model classifies resources according to their response in the park and determines the corresponding adjustment mode for each resource. Confidence gap decision theory is a mathematical framework for dealing with uncertainty optimization problems, particularly suitable for problems where the objective function or constraints contain uncertain parameters. Here, "uncertainty" usually refers to parameters whose exact values ​​cannot be obtained, but whose probability distribution or possible set to which they belong can be obtained.

[0030] Optionally, resource operation parameters refer to the technical parameters of various adjustable resources within the park, which may include, but are not limited to, maximum output limits, minimum output limits (such as the maximum charging and discharging power of batteries, the maximum heating capacity of ground source heat pumps, and the maximum cooling capacity of ground source heat pumps), response speed parameters (such as the time it takes for a resource to complete adjustment from receiving a dispatch command, and the maximum adjustment range of resource output per unit time), energy conversion efficiency (such as the electro-thermal conversion efficiency of thermal storage electric boilers and the electro-cooling conversion efficiency of traditional chillers), and adjustment costs (such as the charging and discharging cost per kWh of energy storage equipment and the adjustment cost of adjustable loads). (Compensation fees); The response status of each resource in the park can be based on the three-level time sequence structure of the electricity spot market ("day-intraday-real-time"), with differentiated requirements for resource response speed and adjustment accuracy. Specifically, the day-ahead market requires resources to provide the next day's 24-hour output plan, with a response requirement of low real-time performance and high stability; the intraday market requires resources to make rolling corrections to the day-ahead plan (e.g., updating every 15 minutes), with a response requirement of medium real-time performance and medium adjustment accuracy; the real-time market requires resources to quickly compensate for photovoltaic output deviations (e.g., minute-level response), with a response requirement of high real-time performance and high adjustment accuracy.

[0031] For example, the basic data corresponding to each resource in the park includes the photovoltaic power generation 96-period forecast curve (divided into 15-minute intervals) and the electricity / heat / cooling load 96-period forecast curve, such as the peak electricity load of the commercial park during the daytime from 10:00 to 18:00, the heat load in winter from 8:00 to 10:00, and the heat load in summer from 12:00 to 8:00. The data includes cooling load, maximum output limit, response speed parameters, energy conversion efficiency, regulation cost, and resource and energy network connectivity. Maximum output limit can include the maximum charging and discharging power of the battery, the maximum heating power of the ground source heat pump (50kW-200kW), and the maximum cooling power. Minimum output limit can include the minimum heating / cooling power of the ground source heat pump. Response speed parameters can be characterized by response rate and response time. The response rate formula is: Response Rate = Maximum Adjustment Range of Rated Output; the response time formula is: Response Time = Adjustment Range per Unit Time Adjustment Rate. Both are normalized (as there may be orders of magnitude differences in adjustment speed due to different resources). Energy conversion efficiency can include the electro-thermal conversion efficiency of thermal storage electric boilers, the electro-cooling conversion efficiency of traditional chillers, and the battery charging efficiency. η charge, battery discharge efficiency η Discharge, ground source heat pump heating efficiency η Heat and the cooling efficiency of ground source heat pumps η Cool; Adjustment costs may include the charging and discharging cost per kWh of energy storage equipment and the adjustment compensation cost of adjustable load; The grid purchase price may be based on the day-ahead / intraday / real-time market time-of-use price; Verification of the connection relationship between resources and energy networks may include connecting batteries to the 380V distribution network, connecting ground source heat pumps to the heating network, and connecting ice storage equipment to the cooling network.

[0032] Information Gap Decision Theory (IGDT) is a method for handling uncertain decision-making problems where accurate probability distribution information is lacking. It is particularly suitable for engineering systems where parameter predictions are biased, but their precise distribution is difficult to obtain. Assume there is an uncertain parameter in the system, whose actual value is... The predicted value of the uncertain parameter is For uncertain sets The construction can be represented as:

[0033] in, It represents the proportion of the prediction error that can be tolerated, also known as the robust radius or uncertainty.

[0034] The IGDT model minimizes the cost function. The goal is to satisfy the performance ceiling under all perturbations. Under the premise of maximizing the uncertainty that the system can tolerate The model is as follows:

[0035] Here, x represents the variable determined by optimization, i.e., the decision variable.

[0036] However, IGDT has the drawback of assuming that the uncertainty is a symmetrical interval, while the uncertainty in real systems is often skewed; at the same time, IGDT lacks a probabilistic basis.

[0037] Confidence Gap Decision Theory (CGDT) inherits the maximum tolerance uncertainty idea from IGDT, but introduces probability distributions or confidence constraints by setting confidence levels. This quantifies the risk of uncertain events, thereby balancing robustness and risk control.

[0038] In this embodiment, the uncertainty is the uncertainty of photovoltaic power output, considering the predicted photovoltaic power output value as... The actual output value of photovoltaic power is The relationship can be represented as:

[0039] in, Indicates robustness parameters; This represents the lower limit of photovoltaic output under the most unfavorable operating conditions of the system.

[0040] Under the constraint of satisfying the confidence level, the maximum robustness parameter can be determined as follows:

[0041] Where P() is the probability function.

[0042] To facilitate the solution, the probabilistic constraints are transformed into a deterministic equivalent form, directly addressing the "worst-case scenario for photovoltaic power output" (i.e., Photovoltaic output is lower than predicted. δ Optimizing the ratio can be expressed as:

[0043] CGDT is introduced into the power balance constraints of the industrial park to handle photovoltaic uncertainties.

[0044] Furthermore, resources can be categorized into three levels: Level 1 resources are allowed to participate in all markets, Level 2 resources participate in day-ahead / intraday markets, and Level 3 resources participate only in real-time markets. The specific categorization rules are as follows: (1) Adjusting speed modeling The resource response speed is determined by the ratio of the resource's response rate to its response time.

[0045] The response rate is expressed as follows:

[0046] in, This indicates the change in power; Indicates the amount of change over time; This indicates the response rate.

[0047] Response time is expressed as follows:

[0048] in, Indicates the response time.

[0049] Because the adjustment speeds of different resources vary by orders of magnitude, response rate and response time are normalized for ease of calculation.

[0050]

[0051]

[0052] in, This represents the normalized response rate; This represents the normalized response time; Represents the response rate among all resources. The minimum value; Represents the response rate among all resources. The maximum value; Indicates the response time among all resources The minimum value; Indicates the response time among all resources The maximum value.

[0053] (2) Market Adaptability Modeling Considering the matching relationship between resource adjustment speed and demand characteristics, define resources. Demand Adaptability indicators , means as follows:

[0054] in, Representing resources i Response time Indicate demand j time scale This represents the tolerance parameter.

[0055] (3) Multi-feature integrated classification method Based on the above models of resource adjustment speed and market adaptability, a multi-feature comprehensive hierarchical parameter based on response speed and market adaptability is proposed. , represented as:

[0056] in, The relative importance weights for response speed; For the response speed, the subscript is used as a characterization term. i Indicates the i-th resource; To adjust the relative importance weight of power; For the first i The regulating power of each resource; The relative importance weight of market adaptability; For the first i The resource in the first j Market adaptability indicators in various scenarios.

[0057] (4) Grading rules To integrate the tiered model with the spot market, thresholds are set for the day-ahead market. Intraday market threshold Real-time market threshold ,in When resources When participating in the market, if Then participation in the day-ahead market is permitted, meaning the participation period is the day-ahead market session; if Then participation in the intraday market is allowed, meaning the participation period is the intraday market session; if This allows participation in the real-time market, meaning the participation period is the real-time market session. For resources i The values ​​of the comprehensive grading parameters in the current market dimension; For resources i The values ​​of the comprehensive grading parameters in the intraday market dimension; For resources i The values ​​of the comprehensive classification parameters in the real-time market dimension.

[0058] Taking participation in real-time market regulation as an example, the indicator function allowing participation in regulation can be expressed as:

[0059] in, Representing resources i Indicator functions that participate in real-time market regulation.

[0060] For example, after obtaining the basic data of the park, the photovoltaic output forecast data is first split into time series (e.g., divided into 96 time periods at 15-minute intervals to form a photovoltaic output forecast time series curve). The load data is then classified and organized into three types of time series data: electrical load, heat load, and cooling load (e.g., peak electrical load data of the commercial park during the day from 10:00 to 18:00, and heat load data from 8:00 to 10:00 in winter). Then, the connection relationship between resources and energy networks is verified (e.g., batteries connected to the 380V power distribution network, ground source heat pumps connected to the heating network) to ensure data integrity and logical consistency. Furthermore, response characteristics can be calculated based on resource operation parameters. For example, the response time of a battery is 2 minutes and the response rate is 50% / min (50% of the maximum adjustable output per minute), the response time of a ground source heat pump is 30 minutes and the response rate is 10% / min, and the response time of an adjustable air conditioning load is 15 minutes and the response rate is 20% / min.

[0061] To match the demand of the electricity market during specific time periods, resources with a response time ≤ 5 minutes and a response rate ≥ 40% / min (such as batteries) are classified as Level 1 resources, allowing participation in all day-ahead, intraday, and real-time markets, with a regulation mode of "real-time tracking deviation compensation." Resources with a response time of 5-30 minutes and a response rate of 15%-40% / min (such as adjustable air conditioning loads) are classified as Level 2 resources, allowing participation in day-ahead and intraday markets, with a regulation mode of "time-based plan correction." Resources with a response time > 30 minutes and a response rate < 15% / min (such as ground source heat pumps) are classified as Level 3 resources, allowing participation only in the day-ahead market, with a regulation mode of "day-ahead output plan locking." The objective of the dispatch model is to "maximize the photovoltaic confidence interval (i.e., the maximum allowable photovoltaic prediction error ratio)" while minimizing the total system operating cost (total operating cost = grid power purchase cost + resource regulation cost - market revenue).

[0062] Furthermore, physical constraints can be established, including but not limited to energy balance constraints, equipment operation constraints, and resource classification constraints. Specifically, energy balance constraints include, for example, the electrical energy balance at a certain time period is defined as "actual photovoltaic output + grid-purchased electricity + battery discharge power = electrical load demand + battery charging power + ground source heat pump power consumption"; equipment operation constraints include, for example, the battery state of charge (SOC) constraint is 20% ≤ SOC ≤ 90%, and the ground source heat pump heating power P constraint is 50kW ≤ P ≤ 200kW; resource classification constraints include, for example, the real-time market regulation power of primary resources is ≤ 80% of their maximum output, and secondary resources are not allowed to participate in real-time market regulation.

[0063] Finally, the objective function, physical constraints, and adjustment variables (such as the real-time adjustment power variable Δ) can be combined. P Resource output variables PresIntegrate these variables to form a complete scheduling model; among them, the adjustment variables need to match the resource level (e.g., level 1 resources correspond to Δ). P 1. Level 2 resources correspond to Δ P 2) Ensure that the model can reflect the logic of "tiered adjustment + uncertainty response".

[0064] S202. Standardize and solve the scheduling model to obtain the resource scheduling results for the park.

[0065] One possible approach is to linearize the scheduling model under nonlinear constraints and then solve the linearized scheduling model to obtain the resource scheduling results corresponding to the park. The resource scheduling results include detailed resource hierarchical scheduling, photovoltaic uncertainty response results, change data of adjustment variables, and satisfaction of operational constraints.

[0066] Nonlinear constraints are non-linear constraints arising from mutually exclusive resource operation states, such as the mutual exclusion of "heating / cooling" in ground source heat pumps (i.e., heating and cooling cannot be performed simultaneously in the same period) and the mutual exclusion of "charging / discharging" in energy storage devices (i.e., charging and discharging cannot be performed simultaneously in the same period). Such constraints cannot be directly expressed by linear equations or inequalities and need to be transformed through mathematical methods. Resource hierarchical scheduling details can record the scheduling behavior data of each resource in the corresponding participating market, including but not limited to time period output value, adjustment trigger conditions, and operation state switching records, which can intuitively reflect the execution of hierarchical strategies. Photovoltaic uncertainty response results are indicators for evaluating the scheduling model's effectiveness in responding to photovoltaic fluctuations, including but not limited to the actual coverage rate of the photovoltaic output confidence interval (the proportion of time periods when the actual photovoltaic output falls within the preset confidence interval), deviation compensation amount (the total adjustment power of resources to correct photovoltaic deviations), and system stability under extreme deviations (such as whether the system can still maintain supply and demand balance when photovoltaic output suddenly drops by 20%).

[0067] For example, the nonlinear term in the scheduling model originates from the operating state of some energy conversion resources. Some resources (such as ground source heat pumps) absorb electrical energy and generate cooling or heating energy; that is, there are two operating states: one is electricity-to-heat, and the other is electricity-to-cooling. These two operating states cannot occur simultaneously. Therefore, two 0-1 variables are needed to represent the operating state of the ground source heat pump, thus generating the nonlinear term. The mathematical model from which the nonlinear term originates is as follows: Mutual exclusion constraints of working modes:

[0068] in, Indicates that the ground source heat pump is in the first i The resource, the first t During certain periods, it is in an electrothermal conversion working state, for example... =1 indicates that the state is in that state. =0 indicates that it is not in that state; Indicates that the ground source heat pump is in the first i The resource, the first t During certain periods, it operates in a state of switching from electric to cold. =1 indicates that the state is in that state. =0 indicates that the state is not in that state.

[0069] Electrothermal confinement:

[0070] in, Indicates that the ground source heat pump is in the first i The resource, the first t Heating output for each time period; Indicates the ground source heat pump i The electro-thermal conversion efficiency of a resource is the proportion of heat energy generated by consuming a unit of electrical energy. Indicates the ground source heat pump i The resource, the first t Input electrical power during each time period.

[0071] Cooling constraint:

[0072] in, Indicates that the ground source heat pump is in the first i The resource, the first t Cooling output for each time period; Indicates the ground source heat pump i The electro-cooling conversion efficiency of a resource is the proportion of cooling energy that can be generated from a unit of electrical energy consumed.

[0073] In this embodiment, the classic Big-M method is used to linearize the nonlinear term. The specific process is as follows: Introducing auxiliary variables , ,

[0074] in, As an auxiliary variable, it is used to linearize the product of the input electrical power and the operating state in the electric-to-thermal state of the ground source heat pump, representing the ground source heat pump's... i The resource, the first t The effective input electrical power components during the electro-thermal conversion period; As an auxiliary variable, it is used to linearize the product of the input electrical power and the operating state in the electric-to-cooling state of the ground source heat pump, representing the ground source heat pump's... i The resource, the first t The effective input electrical power component during the period from electricity to cooling.

[0075] Electro-thermal linearization constraints:

[0076] in, The value of M is a sufficiently large positive number (usually 1.5 times the maximum input electrical power of the ground source heat pump, etc.) used to implement linearization constraints in the Big-M method to ensure the correctness of the logical relationship.

[0077] Electrical-to-cold linearization constraints:

[0078] Furthermore, the CPLEX solver can be used for optimized solutions. The model can be standardized first, such as by setting parameters, for example, setting the maximum number of iterations to 3000, to balance solution efficiency and accuracy, and setting the solution accuracy to [value missing]. To meet the engineering error requirements for the operation of the park's energy system, a computation time limit of 40 minutes was set to avoid excessive computation time due to high model complexity, which could affect the timeliness of scheduling decisions. Furthermore, the standardized scheduling model was imported into CPLEX to initiate iterative computation. In the initial stage of the solution, the uncertainty in the objective function... α The improvement will gradually increase from 5% (e.g., 0.5% per iteration), and the total system operating cost will also be optimized accordingly; when there are 3 consecutive iterations α The change is less than When the cost change is less than 10 yuan, the model is considered to have converged and the calculation is stopped.

[0079] The CPLEX solver can ultimately output key results such as detailed resource hierarchical scheduling (e.g., real-time market output curves for batteries during 96 hours and planned output values ​​for ground source heat pumps during 24 hours), photovoltaic uncertainty handling results (e.g., maximum acceptable photovoltaic prediction error ratio and confidence interval coverage), adjustment variable change data (e.g., real-time adjustment power time series curves), and operational constraint satisfaction (e.g., electricity-heat-cold balance deviation and battery state of charge range), providing quantitative basis for the optimized scheduling of the park's integrated energy system.

[0080] It should be noted that CPLEX, as a mathematical programming solver, has powerful capabilities for solving mixed-integer linear programming and nonlinear programming, and is very suitable for the linearized form of the scheduling model in this embodiment.

[0081] This embodiment integrates photovoltaic uncertainty handling (CGDT) with a resource-level adjustment mechanism to optimize the scheduling of the park's comprehensive energy system. This achieves a leap from extensive unified scheduling to refined hierarchical and collaborative scheduling. Through CGDT, the worst-case fluctuation scenario (maximum α) that the system can withstand is proactively identified, and a scheduling plan is formulated for this scenario. This gives the system inherent immunity to unavoidable photovoltaic fluctuations in actual operation, significantly reducing the risk of power imbalance and achieving a shift from passive response to active defense. The maximum acceptable photovoltaic prediction error ratio α provides operators with a quantitative measure of system robustness. For example, α=20.99% means that the system can still ensure safe operation even when photovoltaic output fluctuates drastically (deviating from the predicted value by ±20.99%), providing a clear safety boundary for decision-making. By classifying resources based on response speed and market adaptability, fast resources (such as batteries) are used for the highest-value real-time deviation compensation and real-time market, while slow resources (such as thermal storage boilers) are used for day-ahead and intraday market energy planning. This avoids resource mismatch by using high-value resources in low-value scenarios, maximizing the value of resources throughout their entire lifecycle. Photovoltaic fluctuations can lead to losses from unplanned power purchases, punitive fees, or frequent resource start-ups and shutdowns. The robust optimization in this embodiment effectively mitigates fluctuations by considering worst-case scenarios and reserving adjustment capacity in advance, reducing unplanned transactions and operational overruns, thereby lowering implicit costs caused by uncertainty. The scheduling model clearly distinguishes the participation capabilities of resources in the day-ahead, intraday, and real-time markets, enabling park operators to formulate bidding and scheduling strategies that precisely match the market structure. The output of this embodiment can include clear resource hierarchical scheduling details, clearly indicating when each resource will be regulated, at what power level, and in which market. This manual-style scheduling instruction greatly improves its executability in actual engineering. Furthermore, by linearizing nonlinear constraints (such as mutually exclusive equipment operation) using methods like the Big-M method, the originally complex and difficult-to-solve nonlinear or stochastic programming problem is transformed into a mature mixed-integer linear programming (MILP) problem.

[0082] The aforementioned park resource scheduling method, based on fundamental data of each resource within the park and confidence gap decision theory, constructs a scheduling model that classifies resources according to their response status and determines the corresponding adjustment mode for each resource, thus determining the degree to which each resource within the park can participate in resource scheduling. Furthermore, the scheduling model is standardized and solved to obtain the corresponding resource scheduling results for the park, achieving perfect resource scheduling. This solution, by introducing a scheduling model that includes resource levels and adjustment modes, maximizes the full lifecycle value of resources, obtains a scheduling strategy that precisely matches the market structure, and ultimately achieves resource scheduling results that are tailored to the specific park.

[0083] Based on the above embodiments, this application provides a detailed explanation of S201. Specifically, this application involves the process of constructing a scheduling model, as follows: Figure 3 As shown, the specific steps include: S301. Based on the basic data of each resource in the park and the confidence gap decision theory, construct the objective function and physical constraints of the scheduling model.

[0084] One feasible approach involves determining the regulation level corresponding to each resource based on resource operating parameters and the response demand of the electricity market at different times; constructing the objective function of the scheduling model based on confidence gap decision theory and the regulation level corresponding to each resource; determining the supply-demand balance relationship and operating characteristics corresponding to each resource based on basic data, and constructing the physical constraints of the scheduling model; and constructing the scheduling model corresponding to the park based on the objective function, physical constraints, and regulation variables.

[0085] Optionally, based on resource operating parameters, determine the response time and adjustment range corresponding to each resource; based on the response demand of the electricity market in each time period, and the response time and adjustment range corresponding to each resource, determine the adjustment level corresponding to each resource.

[0086] Furthermore, based on the adjustment level, the corresponding participation period for each resource is determined.

[0087] Among them, the supply and demand balance relationship refers to the equation relationship between the energy supply and demand in the park at a certain time period. It needs to be constructed separately according to energy type, including electricity supply and demand balance (e.g., supply includes: photovoltaic output + grid purchase + energy storage discharge; demand includes: electricity load + energy storage charging + electricity consumption of energy conversion equipment), heat supply and demand balance (e.g., supply includes: thermal storage electric boiler output + heat pump heating; demand includes: heat load), and cold supply and demand balance (supply includes: ice storage cooling + chiller refrigeration; demand includes: cold load). Resource operation characteristics refer to the inherent technical attributes of various resources during operation, such as the state of charge (SOC) change law of energy storage equipment (SOC rises during charging and falls during discharging), efficiency characteristics of energy conversion equipment (e.g., the efficiency curve of thermal storage electric boiler changes with output), and adjustment delay characteristics of adjustable load (e.g., air conditioning load needs 5 minutes to reach a stable state after adjustment). Participation time period can include but is not limited to day-ahead market time period, intraday market time period, and real-time market time period.

[0088] For example, under the conditions of satisfying the electrical-thermal-cold balance and resource constraints, the objective function for maximizing the system's tolerance range for photovoltaic prediction deviations can be expressed as follows:

[0089] in, The robustness parameter represents the proportion of photovoltaic prediction errors, indicating the system's tolerance for photovoltaic prediction deviations. The objective is to maximize this parameter. δ To broaden the tolerance range; Representing resources i During the scheduling period t electricity ( e ),hot( h ),cold( c The relevant power variables are decision variables in the scheduling model; Representing resources i During the scheduling period t The electrical energy adjustment variable is used to correct power deviation; Indicates the scheduling time period t The amount of electricity purchased from the power grid is one of the decision variables.

[0090] The constraints to be met include power balance constraints, thermal balance constraints, cold balance constraints, day-ahead classification constraints, intraday classification constraints, real-time classification constraints, classification coefficient constraints, CGDT uncertain variable constraints, and equipment operation constraints.

[0091] Optionally, the energy balance constraint can be expressed as:

[0092] in, Indicates the resource number. Indicates the scheduling time period; Indicates the scheduling time period Time Resources The electrical power supplied in the current market; Indicates the scheduling time period Time Resources The electrical power supplied in the intraday market; Indicates the scheduling time period Time Resources The electrical power supplied in the real-time market; Indicates the scheduling time period The predicted output of photovoltaic power; This indicates the demand for electricity load.

[0093] The thermal energy balance constraint can be expressed as:

[0094] in, Representing resources The heat output currently available in the market; Representing resources The heat output provided in the intraday market; Representing resources The thermal power provided in the real-time market; Indicates the scheduling time period t The heat load requirement.

[0095] The cold energy balance constraint can be expressed as:

[0096] in, Representing resources The cooling capacity currently available on the market; Representing resources Cooling power offered in the intraday market; Representing resources The cooling power provided in the real-time market; Indicates the scheduling time period t The cooling load requirement.

[0097] The current hierarchical constraint can be expressed as:

[0098] in, Representing resources i The participation ratio in the day-ahead market, with a value range of [0,1], represents the proportion of the resource's output available in the day-ahead market relative to its maximum capacity; Representing resources i Maximum power output limit; Representing resources i Maximum thermal power output limit; Representing resources i Maximum cooling power output limit.

[0099] Intraday tiered constraints can be expressed as:

[0100] in, Indicates the scheduling time period t Time Resources i Electric power regulation in the intraday market; Indicates the scheduling time period t Time Resources i The amount of heat power adjustment in the intraday market; Indicates the scheduling time period t Time Resources i Cooling power adjustment in the intraday market; Representing resources i The participation ratio in the intraday market, with a value range of [0,1], represents the proportion of the resource's available adjustment capacity in the intraday market relative to its maximum capacity.

[0101] Real-time hierarchical constraints can be expressed as:

[0102] in, Indicates the scheduling time period t Time Resources i Electric power regulation in the real-time market; Indicates the scheduling time period t Time Resources i The amount of thermal power adjustment in the real-time market; Indicates the scheduling time period t Time Resources i Cold power adjustment in the real-time market; Representing resources i The participation ratio in the real-time market, with a value range of [0,1], represents the proportion of the resource's adjustable capacity available in the real-time market to its maximum capacity.

[0103] The tiered coefficient constraint can be expressed as:

[0104] in, Representing resources i The set of grading coefficients (including) , , The value is constrained to be in the interval [0,1] and satisfies the following conditions: This reflects the proportion of resources allocated to different markets.

[0105] The CGDT uncertainty constraint can be expressed as:

[0106] In equipment operation constraints, for energy storage devices, it is necessary to consider their upper and lower limits of energy storage capacity, upper and lower limits of charge / discharge power, and state of charge constraints; for energy conversion devices, it is necessary to consider their upper and lower limits of power and energy conversion efficiency, which can be specifically expressed as follows: 1) Storage battery The charging and discharging process of the battery is modeled as follows:

[0107] in, Indicates that the battery is in The state of charge at any given moment; Indicates that the battery is in The state of charge at any given moment; express Battery charging power at all times; express Battery discharge power at all times; This indicates that the battery has self-damaged; Indicates the battery charging efficiency; Indicates the battery discharge efficiency; This indicates a time interval, typically 1 hour.

[0108] In addition to the above, the battery constraints also need to consider the upper and lower limits of the state of charge, the upper and lower limits of the charge and discharge power, and the state of charge and discharge constraints, which can be expressed as:

[0109]

[0110]

[0111]

[0112] in, Indicates the lower limit of the battery's state of charge; Indicates the upper limit of the battery's state of charge; Indicates the upper limit of charging power; Indicates the upper limit of discharge power; and For 0-1 variables, when When the value is 1, it indicates that the battery is in a charging state; when A value of 1 indicates a discharge state.

[0113] 2) Ice storage Ice storage can convert electrical energy into cold energy, and the equipment has two working modes: refrigeration and ice storage.

[0114] The modeling of the cooling state is as follows:

[0115] in, Indicates the cooling capacity of ice storage; Indicates the number of dual-mode coolers; Indicates the first A dual-mode cooler in The electrical energy absorbed at all times; This indicates the efficiency of the electricity-to-cooling conversion.

[0116] The modeling of the ice-accumulating state is as follows:

[0117]

[0118]

[0119] in, express Ice storage capacity at all times; express Constantly dissipating cooling power; express Ice storage capacity at all times; express Ice storage capacity at all times; This represents the self-loss coefficient of ice storage. express Efficiency of the ice-making and storage process; express Ice-making efficiency at all times; express Constantly reduce cooling efficiency; express The sum of cold energy released by ice in both its cold storage and cold states at all times.

[0120] 3) Thermal storage electric boiler Thermal storage electric boilers can convert electrical energy into heat energy and have energy storage capabilities. Therefore, thermal storage boilers have two working modes: heating and energy storage.

[0121] The modeling for the heating state is as follows:

[0122] in, This indicates the heating capacity of a thermal storage electric boiler; This represents the electrical energy absorbed. This indicates the efficiency of electro-thermal conversion.

[0123] The energy storage state is modeled as follows:

[0124]

[0125]

[0126]

[0127]

[0128] in, This indicates that the thermal storage electric boiler is in The thermal energy stored over time; This indicates that the thermal storage electric boiler is in The thermal energy stored over time; This represents the self-loss coefficient of a thermal storage electric boiler; express Thermal storage capacity; express Constant heat output power; express Real-time thermal storage power; express Constant heat output power; Indicates thermal storage efficiency; Indicates heat release efficiency; This indicates the lower limit of the thermal storage capacity of a thermal storage electric boiler; This indicates the upper limit of the thermal storage capacity of a thermal storage electric boiler; Indicates the upper limit of thermal storage capacity; Indicates the upper limit of heat release power; and For 0-1 variables, when A value of 1 indicates thermal storage; when A value of 1 indicates heat release.

[0129] 4) Ground source heat pump Ground source heat pumps can convert electrical energy into heat or cooling energy. Therefore, ground source heat pumps have two operating modes: cooling and heating. Their constraints are as follows:

[0130]

[0131]

[0132]

[0133] in, This indicates the heat energy released by a ground source heat pump; This indicates the cold energy released by a ground source heat pump; This represents the electrical energy absorbed by the ground source heat pump. Indicates the heating efficiency of a ground source heat pump; Indicates the cooling efficiency of a ground source heat pump; This indicates the lower limit of heat release from a ground source heat pump; This indicates the upper limit of heat output from a ground source heat pump; Indicates the lower limit of cooling; Indicates the upper limit of cooling; For 0-1 variables, when When the value is 1, it indicates that the ground source heat pump is in cooling mode. A value of 0 indicates that the ground source heat pump is in heating mode.

[0134] 5) Traditional water chiller Traditional chiller units mainly consist of chillers, cooling towers, cooling water pumps, and chilled water pumps, consuming electricity for direct cooling. During off-peak electricity consumption periods, the operation of some chiller units can be reduced or stopped, shifting the cooling load to off-peak periods and reducing electricity demand during peak periods. During peak electricity consumption periods, chiller units can be started appropriately, and the number of operating units can be increased to meet the cooling demand of the area, thereby achieving balanced regulation of the cooling load.

[0135] The traditional chiller is modeled as follows:

[0136]

[0137] in, express The capacity of a traditional chiller to store cold energy; express The amount of electricity consumed by traditional water chillers; Indicates the efficiency coefficient; This indicates the upper limit of cold storage capacity for traditional water chillers; This indicates the lower limit of cold storage capacity for traditional chillers.

[0138] S302. Based on the objective function, physical constraints, and adjustment variables, construct the scheduling model corresponding to the park.

[0139] Optionally, based on photovoltaic power output forecast data, load data, resource operation parameters, and a preset photovoltaic forecast error confidence level, the confidence fluctuation range of photovoltaic power output is determined; based on the confidence fluctuation range and adjustment variables, a deviation compensation mechanism is constructed.

[0140] Among them, the photovoltaic output confidence fluctuation range is the output fluctuation range constructed based on photovoltaic output forecast data and a preset confidence level (such as 95%), that is, "photovoltaic forecast value × (1- α Actual photovoltaic output ≤ Photovoltaic forecast value × (1 + α )” α The maximum allowable prediction error ratio is defined by the range width, which reflects the system's tolerance for photovoltaic uncertainties. The deviation compensation mechanism is a dynamic response mechanism that corrects deviations by calling on the adjustment capabilities of different levels of resources when the actual photovoltaic output deviates from the predicted value (exceeding or falling below the confidence fluctuation range). The core is "tiered triggering and precise compensation," that is, small deviations are compensated in real time by first-level resources, medium deviations are compensated by the coordinated efforts of first- and second-level resources, and extreme deviations are compensated by the activation of backup resources (such as emergency diesel generators).

[0141] For example, the objective function aims to maximize the system's ability to withstand uncertainties in photovoltaic output. Specifically, it is defined as maximizing the allowable proportion α of photovoltaic prediction error, which is the maximum range of photovoltaic output fluctuations that the system can withstand under the premise of satisfying all operating constraints. A larger α means that the system is more robust and can cope with more severe photovoltaic fluctuation scenarios while maintaining safe and stable operation.

[0142] By satisfying energy balance constraints (such as electrical energy balance, thermal energy balance, and cold energy balance), resource classification and market participation constraints (such as graded output limits and graded coefficient constraints), equipment operation constraints (such as battery operation constraints, ice storage cooling operation constraints, and ground source heat pump operation constraints), and system operation cost constraints, the scheduling scheme is ensured to be physically feasible and operationally safe.

[0143] Based on the CGDT theory, the set of uncertainties is defined as: Uc(α)={P_pv(t):P_pv_min(t)≤P_pv(t)≤P_pv_max(t)} Where, P_pv_min(t) = P_pv_pred(t)•(1-α), P_pv_max(t) = P_pv_pred(t)•(1+α); Uc(α) is the uncertainty set defined based on CGDT theory, used to characterize the uncertainty range of photovoltaic output, where... α Here are the parameters to be optimized: P_pv(t) represents the actual photovoltaic output at time t; P_pv_min(t) represents the lower limit of photovoltaic output at time t; P_pv_max(t) represents the upper limit of photovoltaic output at time t; P_pv_pred(t) represents the predicted photovoltaic output at time t; and α is the maximum allowable prediction error ratio to be optimized. This interval is centered on the predicted value, and its width is determined by α. The goal of the model is to find the maximum tolerable α while satisfying all constraints.

[0144] Deviation compensation mechanisms are embedded in the constraints of the scheduling model, particularly power balance constraints and real-time market tiering constraints, such as by optimizing real-time flexible adjustment variables (e.g.) , and This allows for dynamic responses to the worst-case scenario of photovoltaic (PV) power output. For example, when solving the scheduling model, it is assumed that the PV power output is at the worst-case point (lower limit) of the confidence interval. Then, this power deficit is compensated by scheduling real-time resources of different levels. In the case of small deviations (corresponding to a small α value), only primary resources (such as fast-responding batteries or some interruptible loads) need to be called to maintain balance. In the case of medium deviations (when α increases, it means that a larger potential deviation needs to be addressed), primary and secondary resources (such as ice storage cooling and thermal storage electric boilers with response speeds in the minute range) are coordinated to work together. In the case of extreme deviations, if the preset backup resources (such as emergency diesel generators) have been included in the scheduling model, when α is maximized, the calling of these tertiary or backup resources will be triggered, and their constraints will be activated to ensure that the system can still maintain balance under the most extreme fluctuations.

[0145] Finally, a robust optimization model was constructed with the objective of maximizing α, constrained by multi-energy flow balance, resource tiering, equipment operation, and cost control, and incorporating an CGDT uncertainty set and a tiered deviation compensation mechanism. After linearization, this model ultimately forms a mixed-integer linear programming model, which can be directly input into commercial solvers such as CPLEX for efficient solving.

[0146] This embodiment abstracts the complex physical system of the industrial park into a structured mathematical model by constructing an objective function and physical constraints, laying a solid data foundation for subsequent optimization. By systematically integrating photovoltaic forecast data, multi-energy load curves, equipment operating parameters, and market information, scheduling decisions are based on a comprehensive and quantitative data foundation, achieving a leap from fuzzy experience to precise calculation. The constructed physical constraints (such as electrical-thermal-cold balance, equipment operating limits, and energy storage dynamic characteristics) are not isolated data points, but rather digital mappings that accurately describe the dynamic coupling relationships of various components within the park. This allows the optimization model to truly reflect the park's operating mechanism, ensuring a high degree of adaptability and physical feasibility between the final scheduling scheme and the actual situation of the park. Based on the aforementioned data foundation, the scheduling model, which integrates CGDT and resource grading, completely changes the traditional scheduling mode. By aiming to maximize the photovoltaic uncertainty tolerance (α), it essentially prepares for a series of possible future scenarios (photovoltaic fluctuations). This makes the generated scheduling scheme inherently robust, able to withstand the impact of uncertainty, and significantly improves the system's operational resilience and power supply reliability. Through the introduced resource grading mechanism, it acts like an intelligent resource scheduler, capable of optimally matching heterogeneous flexible resources within the park in terms of time scale (fast / slow) and spatial dimension (electricity / heat / cold) according to the electricity market's demand for different response speeds (day-ahead, intraday, real-time).This avoids disordered resource allocation or idle capacity, achieving optimal resource utilization and systematically improving the comprehensive utilization efficiency and market value of resources. This solution not only pursues theoretical optimization but also emphasizes practical application. It comprehensively considers operating costs, market returns, and uncertainty risks, aiming for optimal overall system economy throughout the entire scheduling cycle, rather than maximizing the benefits of a single device or part of the system. This helps park operators achieve overall cost reduction and efficiency improvement. By standardizing and linearizing the scheduling model and transforming it into a model that can be efficiently computed by solvers such as CPLEX, high-quality solutions are ensured within a limited time. The final output of the resource hierarchical scheduling details is clear, explicit, and can be directly executed by the Energy Management System (EMS) or by the operator. The precise dispatch instructions from the operators greatly enhance the engineering practicality and implementability of the solution. The "modeling-solving-application" technical path proposed in this solution forms a standardized methodology applicable to various integrated energy systems that include distributed new energy sources and diversified flexible resources. It is not limited to park scenarios but can be extended to regional power grids, microgrids, etc., possessing broad promotion potential and good universality and scalability. The core of the framework is (CGDT robust optimization + resource classification). The model framework is open, and in the future, when the park connects to new resources (such as hydrogen energy, V2G electric vehicles) or faces new market rules, it can quickly adapt to new application requirements by supplementing or modifying the corresponding objective function terms and physical constraints, demonstrating strong technical vitality and adaptability.

[0147] In this embodiment, by first constructing the objective function and physical constraints of the scheduling model, a data foundation is provided for determining a scheduling model suitable for the park; furthermore, by constructing the scheduling model corresponding to the park based on the objective function, physical constraints, and adjustment variables, a tool is provided for obtaining resource scheduling results suitable for the park's situation.

[0148] The above text combined Figures 1 to 3 The park resource scheduling method provided in the embodiments of this application has been described in detail. The apparatus and equipment provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0149] like Figure 4 As shown in the figure, this is a schematic diagram of a park resource scheduling device 600 provided in an embodiment of this application. The park resource scheduling device 600 includes: a model building module 601 and a scheduling module 602, wherein: The model building module 601 is used to build a scheduling model for the park based on the basic data of each resource in the park and the confidence gap decision theory. The scheduling model classifies the resources according to their response status and determines the adjustment mode for each resource. The scheduling module 602 is used to standardize and solve the scheduling model to obtain the resource scheduling results corresponding to the park.

[0150] In one embodiment, the model building module 601 is specifically used for: Based on the basic data of each resource in the park and the confidence gap decision theory, the objective function and physical constraints of the scheduling model are constructed. Based on the objective function, physical constraints, and adjustment variables, a scheduling model corresponding to the park is constructed.

[0151] In one embodiment, the model building module 601 is specifically used for: Based on resource operation parameters and the response demand of the electricity market at different times, the regulation level corresponding to each resource is determined; Based on confidence gap decision theory and the adjustment levels corresponding to each resource, an objective function for the scheduling model is constructed. Based on the basic data, the supply and demand balance relationship and operating characteristics of each resource are determined, and the physical constraints of the scheduling model are constructed. Based on the objective function, physical constraints, and adjustment variables, a scheduling model corresponding to the park is constructed.

[0152] In one embodiment, the model building module 601 is specifically used for: Based on photovoltaic power output forecast data, load data, resource operation parameters, and preset photovoltaic forecast error confidence levels, the confidence fluctuation range of photovoltaic power output is determined. A bias compensation mechanism is constructed based on the confidence fluctuation range and the adjustment variable.

[0153] In one embodiment, the model building module 601 is specifically used for: Based on the resource operation parameters, determine the response time and adjustment range for each resource; Based on the response demand of the electricity market at different times, and the response time and adjustment range of each resource, the adjustment level corresponding to each resource is determined.

[0154] In one embodiment, the model building module 601 is specifically used for: Based on the adjustment level, determine the corresponding participation period for each resource.

[0155] In one embodiment, the scheduling module 602 is specifically used for: The scheduling model is linearized under nonlinear constraints, and the linearized scheduling model is solved to obtain the resource scheduling results corresponding to the park. The resource scheduling results include detailed resource hierarchical scheduling, photovoltaic uncertainty response results, change data of adjustment variables, and satisfaction of operational constraints.

[0156] The park resource scheduling device 600 according to the embodiments of this application can correspondingly execute the method described in the embodiments of this application, and the other operations and / or functions of each module / unit of the park resource scheduling device 600 are respectively for implementing Figure 2 , Figure 3 For the sake of brevity, the corresponding processes of each method in the illustrated embodiments will not be described in detail here.

[0157] This application also provides a computing device. This computing device can be a local computing device or an application server.

[0158] like Figure 5 As shown in the figure, this is a schematic diagram of a computing device provided in an embodiment of this application. The computing device 700 includes a bus 701, a processor 702, a communication interface 703, and a memory 704. The processor 702, the memory 704, and the communication interface 703 communicate with each other via the bus 701.

[0159] The 701 bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0160] The processor 702 can be any one or more of the following processors: central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP), or digital signal processor (DSP).

[0161] Communication interface 703 is used for external communication. For example, communication interface 703 can be used to communicate with terminal 102. Communication interface 703 is used to send resource scheduling results to terminal 102 so that terminal 102 can present the resource scheduling results.

[0162] Memory 704 may include volatile memory, such as random access memory (RAM). Memory 704 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0163] The memory 704 stores executable code, and the processor 702 executes the executable code to perform the aforementioned campus resource scheduling method.

[0164] Specifically, in achieving Figure 4 In the case of the illustrated embodiment, and Figure 4 When the modules or units of the park resource scheduling device described in the embodiment are implemented through software, the execution... Figure 4 The software or program code required for the functions of each module / unit can be partially or entirely stored in memory 704. Processor 702 executes the program code corresponding to each unit stored in memory 704 to execute the aforementioned campus resource scheduling method.

[0165] This application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that a computing device can store, or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium includes instructions that instruct the computing device to execute the aforementioned campus resource scheduling method.

[0166] This application also provides a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this application are generated.

[0167] The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0168] When the computer program product is executed by a computer, the computer executes any of the aforementioned park resource scheduling methods. The computer program product can be a software installation package; when any of the aforementioned park resource scheduling methods needs to be used, the computer program product can be downloaded and executed on the computer.

[0169] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0170] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the scope of protection of this application.

Claims

1. A method for scheduling park resources, characterized in that, The method includes: Based on the basic data of each resource in the park and the confidence gap decision theory, a scheduling model for the park is constructed; wherein, the scheduling model classifies resources according to their response status and determines the adjustment mode for each resource. The scheduling model is standardized and solved to obtain the resource scheduling result corresponding to the park.

2. The method according to claim 1, characterized in that, The scheduling model for the park is constructed based on the basic data and confidence gap decision theory corresponding to each resource in the park, including: Based on the basic data of each resource in the park and the confidence gap decision theory, the objective function and physical constraints of the scheduling model are constructed. Based on the objective function, physical constraints, and adjustment variables, a scheduling model corresponding to the park is constructed.

3. The method according to claim 2, characterized in that, The basic data includes resource types, the connection between resources and energy networks, photovoltaic output forecast data, load data, resource operating parameters, and grid purchase price. The objective function and physical constraints of the scheduling model, constructed based on the basic data corresponding to each resource in the park and the confidence gap decision theory, include: Based on resource operation parameters and the response demand of the electricity market at different times, the regulation level corresponding to each resource is determined; Based on confidence gap decision theory and the adjustment levels corresponding to each resource, an objective function for the scheduling model is constructed. Based on the basic data, the supply and demand balance relationship and operating characteristics of each resource are determined, and the physical constraints of the scheduling model are constructed. Based on the objective function, physical constraints, and adjustment variables, a scheduling model corresponding to the park is constructed.

4. The method according to claim 3, characterized in that, The method further includes: Based on photovoltaic power output forecast data, load data, resource operation parameters, and preset photovoltaic forecast error confidence levels, the confidence fluctuation range of photovoltaic power output is determined. A bias compensation mechanism is constructed based on the confidence fluctuation range and the adjustment variable.

5. The method according to claim 3, characterized in that, The determination of the regulation level corresponding to each resource based on resource operating parameters and the response demand of the electricity market at different times includes: Based on the resource operation parameters, determine the response time and adjustment range for each resource; Based on the response demand of the electricity market at different times, and the response time and adjustment range of each resource, the adjustment level corresponding to each resource is determined.

6. The method according to claim 3, characterized in that, The method further includes: Based on the adjustment level, determine the corresponding participation period for each resource.

7. The method according to claim 1, characterized in that, The process of standardizing and solving the scheduling model to obtain the resource scheduling result corresponding to the park includes: The scheduling model is linearized under nonlinear constraints, and the linearized scheduling model is solved to obtain the resource scheduling results corresponding to the park. The resource scheduling results include detailed resource hierarchical scheduling, photovoltaic uncertainty response results, change data of adjustment variables, and satisfaction of operational constraints.

8. A park resource scheduling device, characterized in that, The device includes: The model building module is used to construct a scheduling model for the park based on the basic data corresponding to each resource in the park and the confidence gap decision theory. The scheduling model classifies the resources according to their response status and determines the adjustment mode for each resource. The scheduling module is used to standardize and solve the scheduling model to obtain the resource scheduling result corresponding to the park.

9. A computing device, characterized in that, Including memory and processor; The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program for performing the method as described in any one of claims 1 to 7.

Images