A dynamic operation domain construction and optimal scheduling method, system and device of a distributed energy system and a storage medium

Abstract

This invention discloses a method, system, device, and storage medium for constructing and optimizing the dynamic operating domain of a distributed energy system. The method includes: collecting incomplete information data of the distributed energy system to construct a polyhedral uncertainty set; establishing a high-dimensional constraint set, using this high-dimensional constraint set as constraints, and optimizing the solution with the objectives of maximizing and minimizing tie-line power; expanding the obtained upper and lower bounds along the time axis to extract the time-series envelope of the dynamic operating domain; constructing a robust optimization model based on the polyhedral uncertainty set, including a day-ahead decision-making stage and an intraday real-time scheduling stage, and embedding the time-series envelope of the dynamic operating domain as a rigid boundary into the feasible domain of the intraday real-time scheduling stage; using a column constraint generation algorithm to iteratively solve the robust optimization model after embedding the rigid boundary, and outputting the optimal unit start-up and shutdown and scheduling strategy. This achieves synergistic optimization of system safety and low carbon emissions.

Description

Technical Field

[0001] This invention relates to a method and system for constructing and optimizing the operation domain, and more particularly to a method, system, device, and storage medium for constructing and optimizing the dynamic operation domain of a distributed energy system. Background Technology

[0002] With the high proportion of distributed new energy access, distributed energy systems face extremely severe uncertainty challenges in short-term operation scheduling. Due to incomplete measurement data, sudden weather changes, etc., it is often impossible to obtain the accurate probability density function of the system's wind and solar power output and multi-energy load, resulting in incomplete information available for decision-making. Existing uncertainty optimization scheduling methods mainly have the following problems: (1) Traditional stochastic programming is highly dependent on full information, and the model is prone to failure. Existing methods mostly assume that the error follows a Gaussian or normal distribution, but under non-full information, the expected cost-minimum strategy calculated based on the error probability distribution is prone to system overload or equipment failure if it encounters extreme non-standard deviations in actual operation. (2) Static feasible domain cannot characterize the time coupling of energy storage and the global pressure of carbon emission space. Most existing feasible domain analyses are static slices for a single moment, ignoring the cross-time energy transfer capability brought about by the state of charge of energy storage equipment, and failing to incorporate the full-day macro carbon emission limit as a boundary condition, resulting in the obtained feasible boundary lacking practical guiding significance. (3) Although traditional two-stage robust optimization can resist extreme scenarios, when encountering extreme source load fluctuations, it may purchase a large amount of high-carbon electricity to maintain system stability, which may lead to a serious over-limit of total carbon emissions, which is contrary to the goal of low-carbon operation. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to solve the problems existing in the prior art and provide a method, system, device and storage medium for constructing and optimizing the dynamic operating domain of a distributed energy system under non-full and uncertain information, so as to achieve synergistic optimization of system security and low carbon emissions.

[0004] Technical solution:

[0005] Firstly, this embodiment provides a method for constructing and optimizing the scheduling of dynamic operating domains in a distributed energy system, including:

[0006] Collect incomplete information data from distributed energy systems and construct a polyhedral uncertainty set based on the incomplete information data;

[0007] A high-dimensional constraint set is established, which includes physical constraints of equipment, cross-time state of charge evolution constraints of energy storage, and macro-carbon emission quota red line constraints. Using this high-dimensional constraint set as the constraint condition, optimization is performed with the goal of maximizing and minimizing tie line power. The obtained maximum and minimum tie line power are set as upper and lower bounds, respectively, and expanded along the time axis to extract the dynamic operating domain time series envelope of the energy system tie line interaction power.

[0008] A robust optimization model is constructed based on the polyhedral uncertainty set, which includes a day-ahead decision-making stage and an intraday real-time scheduling stage. The time-series envelope of the dynamic operating domain is used as a rigid boundary and embedded into the feasible domain of the intraday real-time scheduling stage.

[0009] A column constraint generation algorithm is used to iteratively solve the robust optimization model with embedded rigid boundaries, and output the optimal unit start-up and shutdown and scheduling strategy.

[0010] Furthermore, the polyhedral uncertainty set is constructed by defining a prediction benchmark value, a maximum deviation interval, and introducing a robust conservatism budget parameter, as defined below:

[0011]

[0012]

[0013]

[0014] in, Represents a set of uncertainties in a polyhedron; Indicates the predicted baseline value; ΔL e,t Indicates the maximum deviation; Г e Z is the budget parameter for robustness conservatism. e,t This indicates the direction and extent to which the actual load deviates from the baseline value at time t, where T represents time.

[0015] Furthermore, the optimization solution aimed at maximizing and minimizing tie line power is expressed as follows:

[0016]

[0017]

[0018] Among them, P grid,t* Indicates the power of the tie line, t * For any target time period, the maximum tie-line power obtained by solving is: The minimum tie-line power is The upper and lower boundaries formed by them are Ω represents a high-dimensional constraint set.

[0019] Furthermore, the robust optimization model is a three-layer nested structure of Min-Max-Min, where: the outer Min layer corresponds to the day-ahead decision-making stage and is used to decide on the start-up and shutdown plans of the units; the middle Max layer is used to identify the worst source load fluctuation scenario that leads to the highest system operating cost in the polyhedral uncertainty set; and the inner Min layer corresponds to the intraday real-time scheduling stage and is used to achieve real-time power redistribution through continuous variable optimization under the known day-ahead start-up and shutdown plans and the worst scenario identified by the middle layer.

[0020] Furthermore, the column constraint generation algorithm is used to iteratively solve the robust optimization model embedded with rigid boundaries, including: the main problem corresponding to the day-ahead decision stage receives the worst-case source-load scenario feedback from the sub-problem corresponding to the intraday real-time scheduling stage, solves the optimal day-ahead unit start-up and shutdown plan, and generates a new column of operating variables that follows the dynamic operating domain boundary constraints; after receiving the start-up and shutdown plan solved by the main problem, the sub-problem corresponding to the intraday real-time scheduling stage uses the strong duality theorem to transform the inner continuous optimization model into a maximization problem to search for the worst-case scenario that makes the system operation cost the highest; if the new scenario discovered by the sub-problem will cause the tie-line interaction power to exceed the time-series envelope boundary of the dynamic operating domain, a corresponding cutting plane is generated and fed back to the main problem, so that the main problem adjusts the day-ahead plan in the next iteration until the upper and lower bounds converge.

[0021] Secondly, this embodiment provides a dynamic operating domain construction and optimization scheduling system for a distributed energy system, including:

[0022] Uncertainty modeling module: used to collect incomplete information data of distributed energy systems and construct a polyhedral uncertainty set based on the incomplete information data;

[0023] Dynamic operating domain extraction module: This module is used to establish a high-dimensional constraint set that includes equipment physical constraints, energy storage cross-time period charge state evolution constraints, and macro-carbon emission quota red line constraints. Using this high-dimensional constraint set as the constraint condition, the module optimizes the solution with the goal of maximizing and minimizing tie-line power. The maximum and minimum tie-line power obtained are then set as upper and lower bounds and expanded along the time axis to extract the time-series envelope of the dynamic operating domain of the energy system tie-line interaction power.

[0024] Robust optimization model construction module: used to construct a robust optimization model based on the polyhedral uncertainty set, which includes the day-ahead decision-making stage and the intraday real-time scheduling stage, and to embed the time-series envelope of the dynamic operating domain as a rigid boundary into the feasible domain of the intraday real-time scheduling stage;

[0025] Iterative solution and strategy output module: Used to iteratively solve the robust optimization model with embedded rigid boundaries using a column constraint generation algorithm, and output the optimal unit start-up and shutdown and scheduling strategy.

[0026] Furthermore, the polyhedral uncertainty set specifically includes:

[0027] It is constructed by defining the prediction baseline value, the maximum deviation interval, and introducing a robust conservatism budget parameter, as defined below:

[0028]

[0029]

[0030]

[0031] in, Represents a set of uncertainties in a polyhedron; Indicates the predicted baseline value; ΔL e,t Indicates the maximum deviation; Г e Z is the budget parameter for robustness conservatism. e,t This indicates the direction and extent to which the actual load deviates from the baseline value at time t, where T represents time.

[0032] Furthermore, the dynamic operating domain extraction module optimizes the solution with the objectives of maximizing and minimizing tie-line power as follows:

[0033]

[0034]

[0035] Among them, P grid,t* Indicates the power of the tie line, t * For any target time period, the maximum tie-line power obtained by solving is: The minimum tie-line power is The upper and lower boundaries formed by them are Ω represents a high-dimensional constraint set.

[0036] Furthermore, the robust optimization model is a three-layer nested structure of Min-Max-Min, where: the outer Min layer corresponds to the day-ahead decision-making stage and is used to decide on the start-up and shutdown plans of the units; the middle Max layer is used to identify the worst source load fluctuation scenario that leads to the highest system operating cost in the polyhedral uncertainty set; and the inner Min layer corresponds to the intraday real-time scheduling stage and is used to achieve real-time power redistribution through continuous variable optimization under the known day-ahead start-up and shutdown plans and the worst scenario identified by the middle layer.

[0037] Furthermore, the column constraint generation algorithm is used to iteratively solve the robust optimization model embedded with rigid boundaries, including: the main problem corresponding to the day-ahead decision stage receives the worst-case source-load scenario feedback from the sub-problem corresponding to the intraday real-time scheduling stage, solves the optimal day-ahead unit start-up and shutdown plan, and generates a new column of operating variables that follows the dynamic operating domain boundary constraints; after receiving the start-up and shutdown plan solved by the main problem, the sub-problem corresponding to the intraday real-time scheduling stage uses the strong duality theorem to transform the inner continuous optimization model into a maximization problem to search for the worst-case scenario that makes the system operation cost the highest; if the new scenario discovered by the sub-problem will cause the tie-line interaction power to exceed the time-series envelope boundary of the dynamic operating domain, a corresponding cutting plane is generated and fed back to the main problem, so that the main problem adjusts the day-ahead plan in the next iteration until the upper and lower bounds converge.

[0038] Thirdly, this embodiment provides a computer device including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the dynamic operating domain construction and optimized scheduling method for a distributed energy system.

[0039] Fourthly, this embodiment provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the dynamic operating domain construction and optimized scheduling method for a distributed energy system.

[0040] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0041] This invention constructs a polyhedral uncertainty set based on the maximum deviation interval and conservative budget parameters, thus eliminating the deep dependence of traditional stochastic programming on the full probability density function and effectively improving the survival and defense capabilities of distributed energy systems under incomplete information and extreme operating conditions. By performing extreme value optimization and dimensionality reduction projection on equipment physical constraints, energy storage cross-time coupling states, and the all-day macroscopic carbon quota red line, the dynamic operating domain time-series envelope of tie line power is explicitly extracted, accurately quantifying the system's true spatiotemporal regulation bottom line. A two-stage robust optimization model with nested dynamic operating domain boundaries, namely day-ahead and intraday, is constructed, and the envelope is forcibly implanted as an absolute hard constraint into the real-time scheduling feasible domain. This mechanism completely prevents the risk of carbon emissions exceeding the limit due to the purchase of high-carbon grid power without a bottom line when the system encounters extreme source-load fluctuations, forcing day-ahead decisions to reserve low-carbon units and energy storage in advance. Ultimately, this achieves a rigorous coordination between absolute physical safety and strict low-carbon compliance of the system under uncertainty shocks. Attached Figure Description

[0042] Figure 1 This is a flowchart of the method described in this invention. Detailed Implementation

[0043] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0044] Example 1

[0045] like Figure 1 As shown in the figure, this embodiment provides a method for constructing and optimizing the dynamic operating domain of a distributed energy system, including the following steps:

[0046] S11. Collect incomplete information data of the distributed energy system, and construct a polyhedral uncertainty set based on the incomplete information data;

[0047] Specifically, the prediction data and physical parameters of the distributed energy system are collected, including non-full information data such as prediction errors of wind and solar power output and multi-energy loads, and a polyhedral uncertainty set based on absolute deviation boundary and conservative budget is constructed; the polyhedral uncertainty set does not depend on the probability density function.

[0048] The construction of the polyhedral uncertainty set includes: defining the baseline forecast value and the maximum deviation range for incomplete information data such as prediction errors of wind and solar power output and multi-energy loads, and introducing a robust conservative budget parameter. Taking load fluctuation as an example, its uncertainty set is defined as follows:

[0049]

[0050]

[0051]

[0052] In the formula, Represents a set of uncertainties in a polyhedron; Indicates the predicted baseline value; ΔL e,t Indicates the maximum deviation; Г e Z is a robustness conservative budget parameter used to limit the highest frequency of simultaneous occurrence of extreme adverse fluctuations across the entire control time domain, thereby filtering out extremely low-probability absolutely adverse scenarios; e,t This indicates the direction and extent to which the actual load deviates from the baseline value at time t, where T represents time.

[0053] S12. Establish a high-dimensional constraint set that includes equipment physical constraints, energy storage cross-time state of charge evolution constraints, and macro-carbon emission quota red line constraints. Using this high-dimensional constraint set as the constraint condition, optimize the solution with the goal of maximizing and minimizing tie-line power. Expand the upper and lower bounds obtained from the solution along the time axis and extract the dynamic operating domain time-series envelope of the energy system tie-line interaction power.

[0054] Specifically, the extraction of the dynamic runtime domain temporal envelope includes:

[0055] In the high-dimensional constraint set Ω, the physical capacity upper and lower limits of fuel cells and electrolyzers, the cross-time-period state-of-charge coupling equations for electrochemical energy storage, and the total daily carbon emission quota red line constraint based on explicit analysis are integrated. For any target time period t... * The optimization solutions are performed with the objectives of maximizing and minimizing tie line power, respectively:

[0056]

[0057]

[0058] Among them, P grid,t* Let represent the tie line power, and the maximum tie line power obtained by solving is . The minimum tie-line power is The obtained upper and lower bounds Unfolding along the time axis, an envelope is formed that integrates time coupling characteristics and macroscopic carbon control targets, where Ω represents a set of high-dimensional constraints.

[0059] S13. Construct a robust optimization model based on the polyhedral uncertainty set, which includes the day-ahead decision-making stage and the intraday real-time scheduling stage, and embed the time-series envelope of the dynamic operating domain as a rigid boundary into the feasible domain of the intraday real-time scheduling stage.

[0060] Specifically, the robust optimization model is a three-layer nested structure of Min-Max-Min. The outer layer Min determines the equipment start-up and shutdown plan during the day-ahead decision-making stage; the middle layer Max induces the worst source load fluctuation scenario in the polyhedral uncertainty set; and the inner layer Min performs real-time power reallocation under the given start-up and shutdown plan and the worst scenario.

[0061] Among them, the real-time tie-line power P of the inner layer Min grid,t* It is forcibly truncated within the boundary of the generated dynamic runtime domain, i.e., a hard constraint is imposed:

[0062]

[0063] This forces the decision-making stage to reserve sufficient equipment start-up and shutdown backup and energy storage capacity to resist the risk of exceeding limits caused by extreme fluctuations.

[0064] S14. The column constraint generation algorithm is used to iteratively solve the robust optimization model after embedding rigid boundaries, and the optimal unit start-up and shutdown and scheduling strategy is output.

[0065] Specifically, the iterative solution includes:

[0066] The main problem in the day-ahead decision-making phase receives the worst-case source load scenario feedback from the subproblems, solves the optimal day-ahead unit start-up and shutdown plan, and dynamically generates a new series of operating variables that follow the dynamic operating domain boundary constraints. The subproblems in the intraday real-time scheduling phase, after receiving the start-up and shutdown strategy from the main problem, use the strong duality theorem to transform the inner continuous solution model into a pure maximization problem, identifying the worst-case scenario that results in the highest system operating cost. If the scenario identified by the subproblems causes the system to risk exceeding the dynamic operating domain boundary, a significant penalty will be incurred and fed back to the main problem in the form of a cutting plane, forcing the main problem to modify the day-ahead plan in the next iteration until the upper and lower bounds converge.

[0067] Example 2

[0068] This embodiment takes an electric-hydrogen-thermal coupled distributed energy system as an example. This system interacts with the external main grid via a connection line and is internally equipped with a 150kW photovoltaic power generation device, a 60kW electrolyzer, an 80kW hydrogen fuel cell, a 60kW electric heat pump, and 150kWh electrochemical energy storage. The specific steps are as follows:

[0069] S21: Construct a polyhedral uncertainty set under incomplete information.

[0070] To address the difficulty in obtaining precise probability density information during system operation due to the incomplete forecasting, a maximum forecasting error boundary of ±20% is set for load and photovoltaic data. To avoid over-conservatism, a load robustness conservative budget parameter Γ is introduced. e =6 and photovoltaic robustness conservative budget parameter Г pv =6, which means that within the 24-hour scheduling cycle, a maximum of 6 hours are allowed for extreme and severe operating conditions, such as a 20% load surge or a 20% photovoltaic drop.

[0071] S22: Extract the temporal envelope of the dynamic operating domain under multidimensional constraints.

[0072] First, based on satisfying the physical capacity of each conversion device and the cross-period state of charge evolution equation of energy storage, a macroscopic carbon emission quota constraint of 1000kgCO2 per day is incorporated. Through extreme value optimization, the safe upper and lower bounds of the 24-hour tie-line interactive power are extracted in a dimensionality reduction manner. .

[0073] Next, an operational status analysis was conducted. At 19:00 in the evening, due to the photovoltaic output reaching zero and facing peak evening load, the upper limit of electricity purchase within the traditional physically feasible domain could reach the system's maximum transformer capacity of 250kW. However, in this embodiment, due to the global pressure of the daily carbon emission quota and the decrease in the state of charge (SOC) of the energy storage, the upper limit of the tie-line electricity purchase determined by the extreme value optimization solution is... It was strictly compressed to only 85kW, precisely quantifying the system's true low-carbon regulation baseline at that moment.

[0074] S23: Construct a two-stage robust optimization model for the day-ahead and intraday phases of nested dynamic running domain boundaries.

[0075] First, with the overall goal of minimizing the combined cost of electricity purchase and the cost of carbon penalties, the outer master problem (day-ahead decision stage) determines the 0-1 start-up and shutdown plan for hydrogen fuel cells; the inner sub-problem (intra-day real-time stage) seeks the worst-case fluctuation scenario in the polyhedral uncertainty set and performs power redistribution of continuous variables.

[0076] Secondly, core boundary nesting is performed, requiring that the subproblems, under any adverse scenario, allocate real-time online power P. grid,t* It must be within the boundary of the extracted dynamic operating domain, i.e., P grid,19 ≤85kW, forming a hard safety constraint.

[0077] S24: Use a column constraint generation algorithm for alternating iteration and strategy output.

[0078] The problem is solved by decoupling using a column constraint generation algorithm. In the first iteration, the main problem is that the fuel cell is not turned on due to an information blind spot. The subproblem is then subjected to an extreme load surge scenario of 20% for three consecutive hours from 19:00 to 21:00.

[0079] As shown in Table 1, when the traditional robust optimization method encountered an extreme load surge at 19:00, the system passively purchased a large amount of electricity from the main grid, totaling 145.5 kW, in order to maintain power balance. Since the grid's carbon emission factor was at its peak of 0.8 kg CO2 / kWh at this time, the total carbon emissions for the day rose to 1520.8 kg CO2, severely exceeding the system's carbon red line of 1000 kg CO2.

[0080] Table 1. Comparison of performance indicators of the method of this invention and traditional robust optimization methods in extreme scenarios.

[0081] Optimize control methods Fuel cell start time Maximum online electricity purchase (kW) during evening peak (19:00) Total daily operating cost under extreme scenarios (RMB) <![CDATA[Total carbon emissions (kgCO2) throughout the day under extreme scenarios]]> Carbon emission limit exceeded Traditional two-stage robust optimization 18:00-21:00 145.5 685.4 1520.8 52% over the limit Optimization method of the present invention 15:00-21:00 85.0 712.6 985.4 Meets standards

[0082] In contrast, the method of this invention incorporates hard constraints of the dynamic operating domain within the real-time scheduling space. During optimization, the algorithm is strictly truncated by the upper limit of the dynamic operating domain envelope at 19:00 (85.0 kW). This truncation penalty is fed back to the day-ahead stage through the cutting plane generated by the column constraint algorithm, successfully forcing the scheduling center to advance the fuel cell start-up time to 15:00 for standby during the day-ahead stage and instructing the energy storage to be fully charged ahead of schedule.

[0083] The final results show that the method of the present invention can suppress the total daily carbon emissions under extremely harsh working conditions to 985.4 kgCO2, achieving absolute physical supply and strict low-carbon compliance under the impact of incomplete information, and verifying the excellent effect of the present invention in coordinating physical safety and carbon efficiency compliance under the impact of uncertainty.

[0084] Example 3

[0085] This embodiment provides a dynamic operating domain construction and optimization scheduling system for a distributed energy system, including:

[0086] Uncertainty modeling module: used to collect incomplete information data of distributed energy systems and construct a polyhedral uncertainty set based on the incomplete information data;

[0087] Dynamic operating domain extraction module: It is used to establish a high-dimensional constraint set that includes equipment physical constraints, energy storage cross-time period charge state evolution constraints, and macro carbon emission quota red line constraints. Using this high-dimensional constraint set as the constraint condition, it optimizes the solution with the goal of maximizing and minimizing tie line power. The upper and lower bounds obtained by the solution are expanded along the time axis to extract the dynamic operating domain time series envelope of the energy system tie line interaction power.

[0088] Robust optimization model construction module: used to construct a robust optimization model based on the polyhedral uncertainty set, which includes the day-ahead decision-making stage and the intraday real-time scheduling stage, and to embed the time-series envelope of the dynamic operating domain as a rigid boundary into the feasible domain of the intraday real-time scheduling stage;

[0089] Iterative solution and strategy output module: Used to iteratively solve the robust optimization model with embedded rigid boundaries using a column constraint generation algorithm, and output the optimal unit start-up and shutdown and scheduling strategy.

[0090] The set of uncertainties in the polyhedron is specifically as follows:

[0091] It is constructed by defining the prediction baseline value, the maximum deviation interval, and introducing a robust conservatism budget parameter, as defined below:

[0092]

[0093]

[0094]

[0095] in, Represents a set of uncertainties in a polyhedron; Indicates the predicted baseline value; ΔL e,t Indicates the maximum deviation; Г e Z is a robustness conservative budget parameter used to limit the highest frequency of simultaneous occurrence of extreme fluctuations across the entire control time domain; e,t This indicates the direction and extent to which the actual load deviates from the baseline value at time t, where T represents time.

[0096] The dynamic operating domain extraction module optimizes the solution by maximizing and minimizing tie-line power as follows:

[0097]

[0098]

[0099] Among them, P grid,t* Indicates the power of the tie line, t * For any target time period, the maximum tie-line power obtained by solving is: The minimum tie-line power is The upper and lower boundaries formed by them are Ω represents a high-dimensional constraint set.

[0100] The robust optimization model is a three-layer nested structure of Min-Max-Min, where: the outer Min layer corresponds to the day-ahead decision-making stage and is used to decide on the start-up and shutdown plans of the units; the middle Max layer is used to identify the worst source load fluctuation scenario that leads to the highest system operating cost in the polyhedral uncertainty set; and the inner Min layer corresponds to the intraday real-time scheduling stage and is used to achieve real-time power redistribution through continuous variable optimization under the known day-ahead start-up and shutdown plans and the worst scenario identified by the middle layer.

[0101] The column constraint generation algorithm is used to iteratively solve the robust optimization model embedded with rigid boundaries. This includes: the main problem corresponding to the day-ahead decision-making stage receives the worst-case source-load scenario feedback from the sub-problem corresponding to the intraday real-time scheduling stage, solves the optimal day-ahead unit start-up and shutdown plan, and generates a new column of operating variables that follows the dynamic operating domain boundary constraints; after receiving the start-up and shutdown plan solved by the main problem, the sub-problem corresponding to the intraday real-time scheduling stage uses the strong duality theorem to transform the inner continuous optimization model into a maximization problem to search for the worst-case scenario that makes the system operation cost the highest; if the new scenario discovered by the sub-problem will cause the tie-line interaction power to exceed the time-series envelope boundary of the dynamic operating domain, a corresponding cutting plane is generated and fed back to the main problem, so that the main problem adjusts the day-ahead plan in the next iteration until the upper and lower bounds converge.

[0102] Example 4

[0103] This embodiment provides a computer device, which includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors. When the programs are executed by the processors, they implement the steps of a dynamic operating domain construction and optimized scheduling method for a distributed energy system as described above.

[0104] Example 5

[0105] This embodiment provides a computer storage medium storing a computer program, which, when executed by a processor, implements the steps of a dynamic operating domain construction and optimized scheduling method for a distributed energy system as described above.

[0106] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus (systems), computer devices, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0107] This invention is described with reference to a flowchart of a method according to embodiments of the invention. It should be understood that each step in the flowchart and combinations thereof can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing device, generate instructions for implementing the process. Figure 1 A device for a function specified in one or more processes.

[0108] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 The function specified in one or more processes.

[0109] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 Steps of a specified function in one or more processes.

Claims

1. A method for constructing and optimizing the dynamic operating domain of a distributed energy system, characterized in that, include: Collect incomplete information data from distributed energy systems and construct a polyhedral uncertainty set based on the incomplete information data; A high-dimensional constraint set is established, which includes physical constraints of equipment, cross-time state of charge evolution constraints of energy storage, and macro-carbon emission quota red line constraints. Using this high-dimensional constraint set as the constraint condition, optimization is performed with the goal of maximizing and minimizing tie line power. The obtained maximum and minimum tie line power are set as upper and lower bounds, respectively, and expanded along the time axis to extract the dynamic operating domain time series envelope of the energy system tie line interaction power. A robust optimization model is constructed based on the polyhedral uncertainty set, which includes a day-ahead decision-making stage and an intraday real-time scheduling stage. The time-series envelope of the dynamic operating domain is used as a rigid boundary and embedded into the feasible domain of the intraday real-time scheduling stage. A column constraint generation algorithm is used to iteratively solve the robust optimization model with embedded rigid boundaries, and output the optimal unit start-up and shutdown and scheduling strategy.

2. The method for constructing and optimizing the dynamic operating domain of a distributed energy system according to claim 1, characterized in that, The polyhedral uncertainty set is constructed by defining a prediction benchmark value, a maximum deviation interval, and introducing a robust conservatism budget parameter, as defined below: in, Represents a set of uncertainties in a polyhedron; Indicates the predicted baseline value; ΔL e,t Indicates the maximum deviation; Г e Z is the budget parameter for robustness conservatism. e,t This indicates the direction and extent to which the actual load deviates from the baseline value at time t, where T represents time.

3. The method for constructing and optimizing the dynamic operating domain of a distributed energy system according to claim 1, characterized in that, The optimization solution, which aims to maximize and minimize tie-line power, is expressed as follows: Among them, P grid,t* Indicates the power of the tie line, t * For any target time period, the maximum tie-line power obtained by solving is: The minimum tie-line power is The upper and lower boundaries formed by them are Ω represents a high-dimensional constraint set.

4. The method for constructing and optimizing the dynamic operating domain of a distributed energy system according to claim 1, characterized in that, The robust optimization model is a three-layer nested structure of Min-Max-Min, where: the outer Min layer corresponds to the day-ahead decision-making stage and is used to decide on the start-up and shutdown plans of the units; the middle Max layer is used to identify the worst source load fluctuation scenario that leads to the highest system operating cost in the polyhedral uncertainty set; and the inner Min layer corresponds to the intraday real-time scheduling stage and is used to achieve real-time power redistribution through continuous variable optimization under the known day-ahead start-up and shutdown plans and the worst scenario identified by the middle layer.

5. The method for constructing and optimizing the dynamic operating domain of a distributed energy system according to claim 1, characterized in that, The column constraint generation algorithm is used to iteratively solve the robust optimization model embedded with rigid boundaries. This includes: the main problem corresponding to the day-ahead decision-making stage receives the worst-case source-load scenario feedback from the sub-problem corresponding to the intraday real-time scheduling stage, solves the optimal day-ahead unit start-up and shutdown plan, and generates a new column of operating variables that follows the dynamic operating domain boundary constraints; after receiving the start-up and shutdown plan solved by the main problem, the sub-problem corresponding to the intraday real-time scheduling stage uses the strong duality theorem to transform the inner continuous optimization model into a maximization problem to search for the worst-case scenario that makes the system operation cost the highest; if the new scenario discovered by the sub-problem will cause the tie-line interaction power to exceed the time-series envelope boundary of the dynamic operating domain, a corresponding cutting plane is generated and fed back to the main problem, so that the main problem adjusts the day-ahead plan in the next iteration until the upper and lower bounds converge.

6. A dynamic operating domain construction and optimization scheduling system for a distributed energy system, characterized in that, include: Uncertainty modeling module: used to collect incomplete information data of distributed energy systems and construct a polyhedral uncertainty set based on the incomplete information data; Dynamic operating domain extraction module: This module is used to establish a high-dimensional constraint set that includes equipment physical constraints, energy storage cross-time period charge state evolution constraints, and macro-carbon emission quota red line constraints. Using this high-dimensional constraint set as the constraint condition, the module optimizes the solution with the goal of maximizing and minimizing tie-line power. The maximum and minimum tie-line power obtained are then set as upper and lower bounds and expanded along the time axis to extract the time-series envelope of the dynamic operating domain of the energy system tie-line interaction power. Robust optimization model construction module: used to construct a robust optimization model based on the polyhedral uncertainty set, which includes the day-ahead decision-making stage and the intraday real-time scheduling stage, and to embed the time-series envelope of the dynamic operating domain as a rigid boundary into the feasible domain of the intraday real-time scheduling stage; Iterative solution and strategy output module: Used to iteratively solve the robust optimization model with embedded rigid boundaries using a column constraint generation algorithm, and output the optimal unit start-up and shutdown and scheduling strategy.

7. The dynamic operating domain construction and optimization scheduling system for a distributed energy system according to claim 6, characterized in that, The set of uncertainties in the polyhedron is specifically as follows: It is constructed by defining the prediction baseline value, the maximum deviation interval, and introducing a robust conservatism budget parameter, as defined below: in, Represents a set of uncertainties in a polyhedron; Indicates the predicted baseline value; ΔL e,t Indicates the maximum deviation; Г e Z is the budget parameter for robustness conservatism. e,t This indicates the direction and extent to which the actual load deviates from the baseline value at time t, where T represents time.

8. The dynamic operating domain construction and optimization scheduling system for a distributed energy system according to claim 6, characterized in that, The dynamic operating domain extraction module optimizes the solution by maximizing and minimizing tie-line power as follows: Among them, P grid,t* Indicates the power of the tie line, t * For any target time period, the maximum tie-line power obtained by solving is: The minimum tie-line power is The upper and lower boundaries formed by them are Ω represents a high-dimensional constraint set.

9. A dynamic operating domain construction and optimization scheduling system for a distributed energy system according to claim 6, characterized in that, The robust optimization model is a three-layer nested structure of Min-Max-Min, where: the outer Min layer corresponds to the day-ahead decision-making stage and is used to decide on the start-up and shutdown plans of the units; the middle Max layer is used to identify the worst source load fluctuation scenario that leads to the highest system operating cost in the polyhedral uncertainty set; and the inner Min layer corresponds to the intraday real-time scheduling stage and is used to achieve real-time power redistribution through continuous variable optimization under the known day-ahead start-up and shutdown plans and the worst scenario identified by the middle layer.

10. A dynamic operating domain construction and optimization scheduling system for a distributed energy system according to claim 6, characterized in that, The column constraint generation algorithm is used to iteratively solve the robust optimization model embedded with rigid boundaries. This includes: the main problem corresponding to the day-ahead decision-making stage receives the worst-case source-load scenario feedback from the sub-problem corresponding to the intraday real-time scheduling stage, solves the optimal day-ahead unit start-up and shutdown plan, and generates a new column of operating variables that follows the dynamic operating domain boundary constraints; after receiving the start-up and shutdown plan solved by the main problem, the sub-problem corresponding to the intraday real-time scheduling stage uses the strong duality theorem to transform the inner continuous optimization model into a maximization problem to search for the worst-case scenario that makes the system operation cost the highest; if the new scenario discovered by the sub-problem will cause the tie-line interaction power to exceed the time-series envelope boundary of the dynamic operating domain, a corresponding cutting plane is generated and fed back to the main problem, so that the main problem adjusts the day-ahead plan in the next iteration until the upper and lower bounds converge.

11. A computer device, characterized in that, It includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of a dynamic operating domain construction and optimized scheduling method for a distributed energy system as described in any one of claims 1-5.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of a method for constructing and optimizing the dynamic operating domain of a distributed energy system as described in any one of claims 1-5.

Images