A multi-stage rolling stochastic programming method for regional integrated energy system

Abstract

The application discloses a kind of regional comprehensive energy system multi-stage rolling stochastic programming methods, including according to the development process of regional comprehensive energy system, planning period is divided into several stages, established including pipeline and equipment capacity constraint, power flow constraint, multi-stage planning variable constraint multi-stage planning model, based on information gap decision theory, by weighted summation obtains the uncertain radius of energy system, establishes robust programming and opportunity programming model, carries out regional comprehensive energy system multi-stage stochastic programming;Based on model predictive control method, realize regional comprehensive energy system multi-stage rolling programming, using dynamic programming algorithm in MATLAB simulation platform, obtain rolling stochastic programming scheme.Through the planning method of the application, reduce the uncertainty in the planning process, optimize the energy synergy characteristics, improve the energy synergy comprehensive energy efficiency, reduce the total economic cost, reduce energy consumption, improve energy efficiency, improve the effect of energy saving and emission reduction.

Description

Technical Field

[0001] This invention relates to the field of regional integrated energy planning technology, and in particular to a multi-stage rolling stochastic planning method for regional integrated energy systems. Background Technology

[0002] With the development of society, science, and technology, urbanization has gradually increased worldwide, and the development and utilization of various energy sources have also continuously improved. As a result, Regional Integrated Energy Systems (RIES) have attracted increasing research attention due to their advantages in regional multi-energy complementarity, energy efficiency, and cleanliness, becoming an important research direction for the future development and transformation of energy systems. Compared with traditional single-energy systems, RIES include many different types of energy. Their development is closely related to regional economic development, energy policies, regional resource distribution, people's living standards, and energy consumption habits, and these related factors are constantly expanding, making the energy supply and demand relationship more complex. RIES are gradually evolving into a large-scale nonlinear chaotic system involving multiple aspects. Such a system is easily affected by various internal and external factors, leading to a gradual deviation from the original planning and operation scheme.

[0003] In the process of energy development, unforeseen events are difficult to predict. The development of the current energy system is happening in sync with the development of all aspects of society, and the factors involved will become increasingly numerous.

[0004] Based on the above facts, to ensure that planning schemes better adapt to the development of regional energy demand, it is necessary to make more accurate judgments and estimates on the uncertainties facing energy supply and demand. Furthermore, as an important means of balancing energy supply and demand, planning needs to be adjusted in a timely manner to ensure that demand is met. This requires analyzing the causes and characteristics of uncertainties, and selecting appropriate models and methods to study the mechanisms of these uncertainties in order to clarify their impact on the system. Considering the impact of these multiple uncertainties on planning is one of the keys to further optimizing energy synergy, increasing the proportion of clean energy, and improving overall energy efficiency. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-stage rolling stochastic programming method for regional integrated energy systems, overcoming the shortcomings of existing technologies. This method, based on information gap decision theory and model predictive control, introduces multi-stage planning, information gap decision theory, and model predictive control into the long-term development planning of regional integrated energy systems. It effectively solves the problem of sudden uncertainties that occur during the planning process of regional integrated energy systems and improves the lack of effective connection between the initial planning and later operation of regional integrated energy systems.

[0006] The objective of this invention can be achieved through the following technical solution: a multi-stage rolling stochastic programming method for a regional integrated energy system, characterized by comprising the following steps:

[0007] S1. Divide the regional integrated energy system planning cycle into several stages and establish a multi-stage planning model for the regional integrated energy system. The multi-stage planning model includes energy pipeline and energy equipment capacity constraints, multi-energy flow constraints, and multi-stage planning variable constraints.

[0008] S2. Based on the information gap decision theory, the influence range of renewable energy and load uncertainty in the multi-stage planning process of the energy system is quantified by interval quantification, and the uncertainty radius of the energy system is obtained by weighted summation. Using both risk aversion and risk speculation strategies, combined with the uncertainty radius, multi-stage planning model and constraints, a robust planning and opportunistic planning model is established to carry out multi-stage stochastic planning of the regional integrated energy system.

[0009] S3. Based on the model predictive control method, the renewable energy and load values ​​are predicted in a rolling manner. According to the error between the predicted source and load values ​​and the actual values, the equipment capacity and pipeline type of the planned phase are controlled, and the planning strategy of the phase is changed in a timely manner to reduce the planning deviation caused by the prediction error. The deviation between the actual planning and the prediction is kept within the allowable range, so as to realize the multi-stage rolling planning of the regional integrated energy system.

[0010] S4. Using dynamic programming algorithm on MATLAB simulation platform, the rolling stochastic programming model of regional integrated energy system based on information gap decision theory and model predictive control is solved iteratively to obtain the multi-stage rolling stochastic programming scheme of regional integrated energy system.

[0011] This invention presents a regional integrated energy system planning scheme that considers the uncertainties of long-term planning. To meet the demands of load growth, the energy system planning aims for optimal economic efficiency, maximum revenue, and minimum carbon emissions. A rolling planning stage is incorporated into the planning process to better adapt to unforeseen uncertain events, further optimizing energy synergy characteristics and improving overall energy efficiency. By combining information gap decision theory with model predictive control, multi-level and multi-stage planning is achieved, resulting in reduced energy consumption, improved energy efficiency, and energy conservation and emission reduction.

[0012] As a further aspect of the present invention: the energy equipment includes transformers, wind turbines, energy hubs, and energy conversion equipment; the energy pipelines include power pipelines, natural gas pipelines, and heat pipelines. By utilizing information gap decision theory and model predictive control, the uncertainty in the planning process is reduced. The uncertainty of electricity, natural gas, and heat loads is selected as uncertainty parameters. By utilizing information gap decision theory and model predictive control, the uncertainty in the planning process is reduced, thereby minimizing the overall total economic cost.

[0013] As a further aspect of the present invention: the multi-stage planning model for the regional integrated energy system includes a comprehensive total cost. The total comprehensive cost From stage comprehensive cost Planning investment costs Energy hub revenue and carbon emission reduction costs It consists of four parts:

[0014]

[0015]

[0016] In the formula: This represents the cost of transitioning from state x in stage s to state y in the next stage s+1. The total cost is calculated from... , , and Weighting coefficient composition, It's a carbon tax.

[0017] After optimizing the planning cycle, the uncertainty of various source loads needs to be considered in order to meet the supply and demand of source loads. The decision variables of the multi-stage planning model are the capacity of equipment in newly built and expanded pipelines, substations or energy hubs in each stage.

[0018] As a further aspect of the present invention: the investment cost of the multi-stage planning Including pipeline planning costs Equipment expansion costs Operation and maintenance costs and facility residual value .

[0019] Investment costs cover the entire multi-phase planning process, including the planning costs for new and existing power, gas and heat pipeline expansion, equipment expansion planning and procurement costs, maintenance costs, and facility residual value.

[0020] As a further aspect of the present invention: the energy hub revenue It consists of the electricity load, electricity price, heat load, and unit heat load revenue of typical days in the planning area during winter, summer, and transitional seasons, specifically expressed as follows:

[0021]

[0022] In the formula, and These are the electrical and thermal loads provided to energy hub u, respectively; and These are revenue per unit of electricity load and revenue per unit of heat load, respectively.

[0023] As a further aspect of the present invention: the carbon emission reduction benefits The reduction in carbon tax costs stems from the replacement of coal-fired boilers with centralized heating at energy hubs to provide heat load, and the reduction in carbon emissions due to the use of renewable energy. Specifically:

[0024]

[0025]

[0026]

[0027]

[0028]

[0029] In the formula, , , , These represent the reduction in carbon emissions from heat, electricity, gas, and renewable energy sources in different seasons; The number of days in different seasons; , , These are the energy demands for heat, electricity, and gas on a typical day, where electricity and gas are primary energy inputs, and the secondary heat demand converted from gas and electricity is not included in the primary energy inputs. , , These represent the carbon emissions of heat, electricity, and gas on a typical day in different seasons. and These are the standard coal equivalent conversion factors for heat output and electrical output, respectively; and These are the CO2 emission coefficients for coal and natural gas, respectively. and These represent the carbon emissions from heat before and after the planning.

[0030] As a further aspect of the present invention: the constraints on the energy equipment and energy pipelines are as follows: the constraints of the multi-energy flow model are multi-energy flow constraints; the capacity of the power pipelines corresponding to each pipeline type in stage s, and the capacity of the energy pipelines or energy equipment in the next stage should be greater than or equal to that in the previous stage; the upper limit of the flow of natural gas pipelines and heating pipelines should be greater than the flow calculation results of each pipeline to meet the load demand.

[0031] The constraint condition at stage s is:

[0032]

[0033]

[0034] In the formula, , , These are the upper limits for power flow in power lines, natural gas pipelines, and heating pipelines, respectively. The equipment capacity of substation or energy hub k; The load provided to the energy equipment; i is the i-th pipeline;

[0035] The constraints for the next stage are as follows:

[0036]

[0037] In the formula, and These represent the capacity of the energy pipelines and energy equipment during stage s; and These refer to the capacity of the energy pipelines and energy equipment in the previous stage;

[0038] The specific restrictions on the type of energy equipment and investment cost are as follows:

[0039]

[0040] In the formula, For the construction investment at stage s, For investment constraints; and These represent the wind power generation capacity of stage s and the previous stage of stage s, respectively.

[0041] As a further aspect of the present invention: the multi-stage stochastic planning of the regional integrated energy system based on information gap decision theory specifically involves:

[0042]

[0043] In the formula , , It is the predicted value of electricity, gas, and heat load t during the planning stage; , and These are the uncertain radii of demand for electricity, gas, and heat, respectively.

[0044]

[0045] In the formula, For the uncertain radius of the multi-energy load; , and These are the weighting coefficients for the uncertain radius of multi-energy loads;

[0046] By adopting a risk-avoidance strategy and combining the uncertain radius, multi-stage programming model, and constraints, a robust stochastic programming model is established as follows:

[0047]

[0048] Using a risk-speculation strategy, and combining the uncertainty radius, multi-stage programming model, and constraints, a chance stochastic programming model is established as follows:

[0049]

[0050] Risk aversion coefficient; This represents the risk and speculation coefficient. for The overall total cost when the uncertain parameter takes a definite value; This represents the total number of stages.

[0051] As a further aspect of the present invention: the step of realizing the multi-stage rolling planning of the regional integrated energy system in S3 includes:

[0052] S31. The predictive model is used to describe the controlled object. It can predict the future state of the controlled object. Based on the load and planning scheme of the system at stage t, the output of stage t+1 can be predicted. The basic linear state-space expression is as follows:

[0053]

[0054] In the formula, X(t) represents the load that the system can bear under the current planning scheme; Y p (t) represents the system's predicted load; This represents the change in the planning scheme, that is, the increment of the planning scheme in the next stage; system D, E, and F are the system matrix, input matrix, and output matrix, respectively;

[0055] S32. Rolling optimization: In actual operation, due to the uncertainty of the system, it is necessary to correct the overall planning scheme of each stage. In each stage, based on the optimization performance index of that stage, the optimal planning scheme for the next stage is solved. The objective function of tracking performance aims to reduce the difference between the planning and the actual load caused by uncertainty, as shown below.

[0056]

[0057]

[0058]

[0059]

[0060] In the formula, This indicates the system's predicted load; Indicates the actual load of the system; This represents the comprehensive deviation index between the predicted loads at time t; The comprehensive deviation index between the predicted load and the actual required power supply load at time t. , , , , , and These represent the deviations and distribution coefficients between the predicted loads of electricity, natural gas, and heat at time t and the actual required power supply loads. , , This represents the actual supply and demand load of electricity, natural gas, and heat at time t;

[0061] S33. MPC feedback correction is used to modify the stage planning, reducing system uncertainty. Compensation planning is performed at each stage based on the deviation of the objective function to improve system robustness. The feedback correction formula is as follows:

[0062]

[0063] In the formula, e represents the systematic error.

[0064] As a further aspect of the present invention: the dynamic programming algorithm in S4 specifically involves: progressively determining the minimum cost from the final state to the initial point in a multi-stage planning process, and calculating the cost using the following recursive equation:

[0065]

[0066] In the formula, This represents the cost of transitioning from state x in stage s to state y in the next stage s+1. and Let x be the planning cost for state x in stage s and y be the planning cost for state y in stage s+1, respectively.

[0067] The beneficial effects of this invention are:

[0068] 1. By applying the planning scheme of this invention, a multi-stage plan for a regional integrated energy system can be established, taking into account the uncertainties of long-term planning. This plan includes factors such as investment and construction costs, operating costs, energy hub revenue, and carbon emission reduction. It focuses on load uncertainty, selects the uncertainties of electricity, natural gas, and heat loads as uncertainty parameters, and utilizes information gap decision theory and model predictive control to reduce sudden uncertainties in the planning process and minimize the overall economic cost.

[0069] 2. This invention incorporates a rolling planning step into the planning process, which is more adaptable to sudden and uncertain events compared to deterministic single-stage planning. By controlling the prediction and actual errors within allowable ranges and adjusting the planning scheme in a rolling manner, the energy synergy characteristics are further optimized, and the overall energy efficiency is improved.

[0070] 3. This invention combines information gap decision theory with model predictive control. The former optimizes economic costs, while the latter optimizes the error between the planned scheme and actual needs, enabling multi-level and multi-stage planning. This results in reduced energy consumption, improved energy efficiency, energy conservation and emission reduction, and improved the effective connection between the initial planning and later operation of the regional integrated energy system. Attached Figure Description

[0071] Figure 1 This is a flowchart illustrating a multi-stage rolling stochastic programming method for a regional integrated energy system according to the present invention.

[0072] Figure 2 This is a schematic diagram of the solution process for a multi-stage rolling stochastic programming method for a regional integrated energy system according to the present invention.

[0073] Figure 3 This is a schematic diagram of a multi-stage rolling stochastic programming process for a regional integrated energy system according to the present invention;

[0074] Figure 4 This is a schematic diagram of the energy equipment and energy pipeline planning in the Tianjin Beichen Demonstration Zone, as an embodiment of the present invention.

[0075] Figure 5 This is a comparison chart of the errors between multi-stage rolling stochastic programming and multi-stage programming in this invention;

[0076] Figure 6 This is a cost comparison chart for each stage of the multi-stage rolling stochastic programming of this invention. Detailed Implementation

[0077] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar symbols denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0078] like Figure 1-6As shown, this invention discloses a multi-stage rolling stochastic programming method for a regional integrated energy system. The method involves dividing the planning cycle into several stages based on the actual development process of the regional integrated energy system. To meet the demands of load growth, the energy system planning aims for optimal economy, maximum revenue, and minimum carbon emissions. A multi-stage planning model for the regional integrated energy system is established, including constraints on energy pipelines, energy equipment capacity, multi-energy flow constraints, and multi-stage planning variables. Based on information gap decision theory, the influence range of renewable energy and load uncertainties during the planning process is quantified using interval quantification. The uncertainty radius of the energy system is obtained through weighted summation. Then, using both risk aversion and risk speculation strategies, combined with the uncertainty radius, the multi-stage planning model, and constraints, robust programming and opportunistic programming models are established for multi-stage stochastic programming of the regional integrated energy system. During the planning process, renewable energy and load values ​​are predicted on a rolling basis using model predictive control methods. Based on the error between the predicted and actual source load values, the equipment capacity and pipeline type planned for this stage are controlled, and the planning strategy for this stage is changed in a timely manner to reduce the planning deviation caused by the prediction error, so that the deviation between the actual planning and the prediction is within the allowable range, thereby realizing the multi-stage rolling planning of the energy system. The dynamic programming algorithm is used on the MATLAB simulation platform to solve the rolling stochastic programming model of the regional integrated energy system based on information gap decision theory and model predictive control, and obtain the multi-stage rolling stochastic programming scheme of the energy system.

[0079] In the stage of establishing a multi-stage planning model for the basic energy system, after optimizing the planning cycle before planning, it is necessary to consider the uncertainty of various energy loads and carry out stochastic planning for each stage, including planning the energy loads of substations, energy stations, distributed energy and various energy pipelines, in order to meet the supply and demand of the source loads. The decision variables of the multi-stage planning model are the new and expanded pipelines in each stage, as well as the capacity of equipment in substations or energy hubs.

[0080] Total cost of multi-stage planning model From stage comprehensive cost The overall cost of each stage consists of planned investment costs. Energy hub revenue Carbon emission reduction costs It consists of four parts.

[0081]

[0082]

[0083] In the formula: This represents the cost of transitioning from state x in stage s to state y in the next stage s+1. The total cost is calculated from... , , and Composition of weighting coefficients. It's a carbon tax.

[0084] Investment costs of multi-stage planning It mainly includes four parts: pipeline planning costs Equipment expansion costs Operation and maintenance costs residual value of facilities .

[0085] Part 1: Pipeline Planning Costs This includes the planning costs for new construction and expansion of existing power, gas, and heating pipelines. and the corresponding construction costs Specifically, it is expressed as follows:

[0086]

[0087]

[0088] In the formula, This represents the planned costs for new and existing power, natural gas, and heat pipeline expansions. Represents the procurement cost of equipment expansion planning. Represents maintenance costs, Represents surplus value. Represents the present value factor. represents the discount rate, and t represents the number of years from the start of the plan;

[0089] Part Two: Equipment Expansion Costs Including purchase cost Installation costs of transformers, wind turbines, and energy conversion equipment in energy hubs Specifically, it is expressed as follows:

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098] In the formula, This represents the planned costs for new and existing power, natural gas, and heat pipeline expansions. This represents the corresponding construction cost. Represents the procurement cost of equipment expansion planning. This indicates the cost of installing equipment, including transformers, energy conversion equipment in energy hubs, and wind turbines. Represents maintenance costs, Represents surplus value. Represents the present value factor. represents the discount rate, and t represents the number of years from the start of the plan. This indicates the unit construction cost of the pipeline. Let i be the length of the i-th pipeline; and These refer to the type and capacity of pipelines and equipment during stage s; and These are the unit costs of pipelines and equipment, respectively. For the expansion increment of the device; and These are the equipment installation costs and land purchase costs, respectively.

[0099] Part Three: Operation and Maintenance Costs of Pipelines and Equipment Specifically, it is expressed as follows:

[0100]

[0101] In the formula, and These are the maintenance factors for pipelines and equipment, respectively. For operation and maintenance time;

[0102] Part Four: Residual Value of Pipelines and Equipment at the End of the Planning Period Specifically, it is expressed as follows:

[0103]

[0104]

[0105]

[0106] In the formula, Net residual value rate; It refers to the lifespan of the equipment. and For depreciation expenses; This refers to the depreciation period.

[0107] Furthermore, energy hub revenue It consists of the electricity load, electricity price, heat load, and unit heat load revenue of typical days in the planning area during winter, summer, and transitional seasons, as detailed below:

[0108]

[0109] In the formula, and These are the electrical and thermal loads provided to the energy hub u, respectively. and These are revenue per unit of electricity load and revenue per unit of heat load, respectively.

[0110] Furthermore, the benefits of carbon emission reduction come from the replacement of coal-fired boilers with centralized heating in energy hubs to provide heat load, and from the reduction in carbon emissions due to the use of renewable energy, which in turn leads to a reduction in carbon tax costs, as detailed below:

[0111]

[0112]

[0113]

[0114]

[0115]

[0116] In the formula, , , , These represent the reduction in carbon emissions from heat, electricity, gas, and renewable energy sources in different seasons; For the number of days in different seasons; , , These are the energy demands for heat, electricity, and gas on a typical day, where electricity and gas are primary energy inputs, and the secondary heat demand converted from gas and electricity is not included in the primary energy inputs. , , These represent the carbon emissions of heat, electricity, and gas on a typical day in different seasons. and These are the standard coal equivalent conversion factors for heat output and electrical output, respectively; and These are the CO2 emission coefficients for coal and natural gas, respectively. and These represent the carbon emissions from heat before and after the planning.

[0117] Furthermore, the constraints on the energy equipment and energy pipelines are constraints of the multi-energy flow model. Specifically, the constraints on the energy equipment and energy pipelines are: the constraints of the multi-energy flow model are multi-energy flow constraints; in addition, at stage s, the upper limit of the power, gas, and heat flow corresponding to each pipeline type should be greater than the flow calculation result of each pipeline meeting the load demand; at the next stage, the capacity of the pipeline or equipment should be greater than or equal to that of the previous stage.

[0118] The constraints at stage s are:

[0119]

[0120]

[0121] In the formula, , , These are the upper limits for power flow in the electricity, gas, and heat pipelines, respectively. The equipment capacity of substation or energy hub k; The load provided to the energy equipment; i is the i-th pipeline;

[0122] The constraints for the next stage are:

[0123]

[0124] In the formula, and These represent the capacity of the energy pipelines and energy equipment during stage s; and These refer to the capacity of the energy pipelines and energy equipment in the previous stage.

[0125] Furthermore, the specific restrictions on equipment type and investment cost are as follows:

[0126]

[0127] In the formula, For the construction investment at stage s, For investment constraints; and These represent the wind power generation capacity of stage s and the previous stage of stage s, respectively.

[0128] Furthermore, multi-stage stochastic programming based on information gap decision theory addresses the risk-averse strategies typically employed by conservative decision-makers in uncertain environments. These strategies aim to maximize the adverse effects of the uncertainty radius while ensuring the optimization objective remains within an acceptable range, thereby guaranteeing the achievement of the minimum expected objective. In this invention, we propose that a larger uncertainty radius leads to lower sensitivity to uncertainty fluctuations, better model robustness, and stronger risk aversion capabilities. Therefore, we establish the mathematical model corresponding to this strategy:

[0129]

[0130] In the formula Let be the objective function value of the above equation. This represents the risk aversion bias factor, indicating that the expected cost is higher than the expected cost. The degree of deviation.

[0131] Another risk-speculation strategy seeks to optimize the target value as much as possible; that is, the smaller the radius of uncertainty, the greater the chance of achieving the expected optimization goal. However, the more sensitive the solution is to fluctuations in uncertainty, the greater the risk the system faces. The corresponding mathematical model is:

[0132]

[0133] In the formula The risk-speculative deviation factor represents the expected cost being lower than the actual cost. The degree of deviation.

[0134] Since this invention also focuses on load uncertainty, the uncertainties of power lines, natural gas pipelines, and heating pipelines are selected as uncertainty parameters, and their fluctuation range is expressed as follows:

[0135]

[0136] In the formula , , and This represents the predicted values ​​of electricity, gas, and heat loads at planning time t. , , The radii representing the electricity, gas, and heat load demands are respectively uncertain. The overall uncertainty radius of the system is obtained by weighted summation:

[0137]

[0138] In the formula , and The weighting coefficient representing the uncertain radius of multi-energy load demand.

[0139] Then, a risk-averse strategy programming model is introduced into the stochastic programming model to obtain a multi-stage stochastic programming model for regional integrated energy based on information gap decision theory, as shown below:

[0140] By adopting a risk-avoidance strategy and combining the uncertain radius, multi-stage programming model, and constraints, a robust stochastic programming model is established as follows:

[0141]

[0142] Using a risk-speculation strategy, and combining the uncertainty radius, multi-stage programming model, and constraints, a chance stochastic programming model is established as follows:

[0143]

[0144] In the formula, Risk aversion coefficient; This represents the risk and speculation coefficient. For when That is, the total cost when the uncertain parameter takes a certain value; This represents the total number of stages.

[0145] Furthermore, based on the model predictive control method, multi-stage rolling planning incorporates a rolling planning stage into the multi-stage stochastic planning process, establishing a rolling optimization model for the model predictive control method. This minimizes the prediction and actual errors, promptly addresses fluctuations in multi-energy load demand due to regional development, and reduces the deviation between prediction and actual demand during the regional integrated energy system planning process. Since the main objective is to adjust the planning scheme based on predicted and actual load information, the prediction and actual errors are controlled within acceptable limits. This includes the following steps:

[0146] First, a predictive model is established to describe the controlled object. This model can predict the future state of the controlled object, and the output of stage t+1 can be predicted based on the load and planning scheme at stage t. The basic linear state-space expression is as follows:

[0147]

[0148] In the formula, X(t) represents the load that the system can bear under the current planning scheme; Y p (t) represents the system's predicted load; This represents the change in the planning scheme, that is, the increment of the planning scheme in the next stage; system D, E, and F are the system matrix, input matrix, and output matrix, respectively.

[0149] Next, rolling optimization is performed. In actual operation, due to the uncertainties inherent in the system, it is necessary to correct the overall planning scheme for each stage. At each stage, based on the performance indicators optimized for that stage, the optimal planning scheme for the next stage is determined. The objective function for tracking performance aims to reduce the difference between the planned and actual loads caused by uncertainties, as shown below.

[0150]

[0151]

[0152]

[0153]

[0154] In the formula, This indicates the system's predicted load; Indicates the actual load of the system; This represents the comprehensive deviation index between the predicted loads at time t; The comprehensive deviation index between the predicted load and the actual required power supply load at time t. , , , , , and These represent the deviations and distribution coefficients between the predicted loads of electricity, natural gas, and heat at time t and the actual required power supply loads. , , Let t represent the actual supply and demand load of electricity, natural gas, and heat at time t.

[0155] Then, feedback correction is performed. MPC uses feedback correction to modify the stage plan, reducing system uncertainty. The principle of MPC is to make adjustments based on the prediction domain (t0 to t0+t1+…+t) in stage t0. n The multi-energy load forecasting information is used to obtain multi-stage planning within the forecast domain through optimization. Only the planning within the first stage is executed. In the next planning cycle, the forecast domain is shifted forward by one stage, and based on the latest system state, plans from t1 to t1+t2+…+t are executed. n The time domain is used for prediction and optimization, and the planning scheme for the second stage is executed. This process is repeated continuously, with the prediction domain being compressed towards the end of the planning cycle and the control domain being shifted backward, until planning schemes for all time periods of the planning cycle are generated.

[0156] At each stage, compensation planning is performed based on the deviation of the objective function to improve system robustness. The feedback correction formula is as follows:

[0157]

[0158] In the formula, e represents the systematic error.

[0159] Furthermore, based on the information gap decision theory, the regional integrated energy rolling planning model is solved using a dynamic programming algorithm. The minimum cost from the final state to the initial point is determined step-by-step through the multi-stage planning process, and the cost is calculated using the following recursive equation:

[0160]

[0161] In the formula, This represents the cost of transitioning from state x in stage s to state y in the next stage s+1; and Let x be the planning cost for state x in stage s and y be the planning cost for state y in stage s+1, respectively.

[0162] The following analysis uses an example to illustrate a multi-stage stochastic rolling planning scheme for a regional integrated energy system:

[0163] This example selects a typical northern urban area located in Beichen District, Tianjin, China. This region has a warm temperate semi-humid monsoon climate with significant seasonal temperature variations. The average temperature in autumn and winter is approximately -0.8℃, thus requiring centralized heating. The planning area is an approximately rectangular region, 6.35 km long and 3.46 km wide, covering an area of ​​approximately 22.23 square kilometers, with a permanent population of about 60,000. The planning area is a rapidly developing industrial-urban interaction zone in northern China, experiencing rapid growth in energy demand and possessing enormous energy utilization potential. Many areas within this urban area are currently under planning and construction, some of which have new multi-energy load demands. Figure 4 It describes the regional development plan for the Beichen District of Tianjin by 2037.

[0164] Currently, the area has three 110kV substations (FDY, XFT, and FHB), one 220kV substation (RHY), one energy hub (EH1), and one natural gas valve station for energy supply. Combined heat and power (CHP) is the core energy supply equipment of the energy hub. The 220kV RHY substation supplies power to the 110kV substations in the area from external sources. Currently, 31 electrical loads in the area are supplied by 28 10kV lines. The 29 gas loads within the demonstration area are supplied by a gas gate station using 37 gas pipelines. One of these pipelines is supplied by the CHP as the core energy hub, providing heat to some industrial users and heating residential buildings through gas heat exchange.

[0165] This embodiment utilizes the proposed multi-stage planning method, considering four factors: construction investment cost, operating cost, energy hub revenue, and carbon emission reduction revenue. A dynamic programming algorithm is employed on the MATLAB simulation platform to iteratively solve a rolling stochastic programming model for a regional integrated energy system based on information gap decision theory and model predictive control, obtaining a rolling stochastic programming scheme. Finally, the model's effectiveness is verified using a typical northern urban area in Beichen District, Tianjin, China.

[0166] Table 1 Rolling Stochastic Programming Scheme for Regional Integrated Energy System Based on Information Decision Theory and Model Predictive Control

[0167]

[0168] Table 1 shows the details of the planning scheme after rolling stochastic programming of the regional integrated energy system based on information gap decision theory and model predictive control.

[0169] Figure 5 This study investigates the error distribution between predicted and actual loads for three methods of regional integrated energy systems based on information gap decision theory and model predictive control: rolling stochastic programming, rolling programming, and multi-stage programming at different times. Figure 5 It can be clearly demonstrated that the rolling stochastic programming of the regional integrated energy system based on the information gap decision theory and model predictive control in this invention can continuously track errors and adjust the planning in a timely manner.

[0170] Figure 6 (a) shows the change in the average error between planned energy supply and load at each stage. In the initial planning phase, load growth slowed, leading to a significant deviation from load forecasts. Therefore, with planned construction exceeding actual load, some energy facilities were redundantly planned, resulting in an increased average error in stage 2. To reduce this redundancy, although there was almost no further planning or construction in stages 2 and 3, the average deviation between energy supply and load gradually decreased from stages 3 to 5 as load growth recovered and rolling and stochastic planning were implemented, thus achieving a balance between supply and demand deviations from nearly 10% to around 5%. In contrast, Figure 6 (b) shows the distribution of total investment cost, pipeline investment cost, and equipment investment cost for each stage. Clearly, to reduce pipeline and equipment redundancy caused by slowing or even halting load growth, and given the small planned construction amounts for stages 2 and 3, investment costs are primarily concentrated in stages 4 and 5. Therefore, the sum of investment costs in stages 4 and 5 accounts for approximately 77% of the total planned cost. Specifically, stage 4 involves the construction of Energy Station EH2 (No. 2), while pipeline construction and expansion plans are distributed across stages 1, 4, and 5, increasing with load growth in stage 4. Figure 6 (c) shows the total carbon emission reduction at each stage of the energy station and the individual carbon emission reduction of energy station EH1 and energy station EH2. Figure 6 (d) provides the total revenue for each stage of the energy station, as well as the individual revenue figures for energy stations EH1 and EH2. Figure 6 (c) and (d) are clearly related. In the later stages of the plan, phases 4 and 5, carbon emission reduction and energy station revenue increased significantly due to the construction of the EH2 energy station. The load provided by EH2 also grew rapidly, accounting for the majority of carbon emission reduction and energy station revenue in the later stages of the plan. Therefore, compared with phase 4, the overall carbon emission reduction and energy station revenue in phase 5 increased by approximately two times.

[0171] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-stage rolling stochastic programming method for a regional integrated energy system, characterized in that: Includes the following steps: S1. Divide the regional integrated energy system planning cycle into several stages and establish a multi-stage planning model for the regional integrated energy system. The multi-stage planning model includes energy pipeline and energy equipment capacity constraints, multi-energy flow constraints, and multi-stage planning variable constraints. S2. Based on the information gap decision theory, conduct multi-stage stochastic programming for a regional integrated energy system, specifically as follows: In the formula , , These are the predicted values ​​of electricity, natural gas, and heat loads at time t during the planning phase; , and These are the uncertain radii of demand for electricity, natural gas, and heat, respectively. In the formula, For the uncertain radius of the multi-energy load; , and These are the weighting coefficients for the uncertain radius of multi-energy loads; By adopting a risk-avoidance strategy and combining the uncertain radius, multi-stage programming model, and constraints, a robust stochastic programming model is established as follows: Using a risk-speculation strategy, and combining the uncertainty radius, multi-stage programming model, and constraints, a chance stochastic programming model is established as follows: Risk aversion coefficient; This represents the risk and speculation coefficient. for The overall total cost when the uncertain parameter takes a definite value; This represents the total number of stages; Total cost; Let be the stage comprehensive cost, representing the cost of transitioning from state x in stage s to state y in the next stage s+1; S3. Based on model predictive control, realize multi-stage rolling planning of regional integrated energy system, including: S31. Based on the system load at time t and the planning scheme, predict the output at time t+1. The basic linear state-space expression is as follows: In the formula, W(t) represents the load that the system can bear at time t under the current planning scheme; Y p (t) represents the predicted load of the system at time t; This represents the change in the planning scheme, that is, the increment of the planning scheme at time t in the next stage; D, E, and F are the system matrix, input matrix, and output matrix, respectively. S32. At each stage, based on the performance index of that stage, solve for the optimal planning scheme for the next stage from that stage. The objective function of tracking performance aims to reduce the difference between the planned and actual loads caused by uncertainty, as shown below. In the formula, This represents the actual load on the system at time t. The comprehensive deviation index between the predicted load and the actual required power supply load at time t. , , , , ,and These represent the deviations and distribution coefficients between the predicted loads of electricity, natural gas, and heat at time t and the actual required power supply loads. , , This represents the actual supply and demand load of electricity, natural gas, and heat at time t; S33. The stage plan is corrected through MPC feedback correction. The feedback correction formula is as follows: In the formula, e represents the systematic error; S4: Solving the regional integrated energy rolling planning model based on information gap decision theory, obtaining a multi-stage rolling stochastic planning scheme for the regional integrated energy system. The dynamic programming algorithm is used to solve the problem, progressively determining the minimum cost from the final state to the initial point in the multi-stage planning. The cost is calculated using the following recursive equation: In the formula, and Let x be the planning cost for state x in stage s and y be the planning cost for state y in stage s+1, respectively.

2. The multi-stage rolling stochastic programming method for a regional integrated energy system according to claim 1, characterized in that: The energy equipment includes transformers, wind turbines, energy hubs, and energy conversion equipment; The energy pipelines include power lines, natural gas pipelines, and heating pipelines.

3. The multi-stage rolling stochastic programming method for a regional integrated energy system according to claim 1, characterized in that: The multi-stage planning model for the regional integrated energy system includes a comprehensive total cost. The total comprehensive cost From stage comprehensive cost Composition, the overall cost of the aforementioned stages Investment costs from multi-stage planning Energy hub revenue Carbon emission reduction costs Four parts composition: In the formula: , and These are weighting coefficients. It's a carbon tax.

4. The multi-stage rolling stochastic programming method for a regional integrated energy system according to claim 3, characterized in that: The investment cost of the multi-stage planning Including pipeline planning costs Equipment expansion costs Operation and maintenance costs and facility residual value .

5. The multi-stage rolling stochastic programming method for a regional integrated energy system according to claim 3, characterized in that: The energy hub revenue It consists of the electricity load, electricity price, heat load, and unit heat load revenue of typical days in the planning area during winter, summer, and transitional seasons, and is expressed as: In the formula, and These are the electrical and thermal loads provided to the energy hub u, respectively. and These are revenue per unit of electricity load and revenue per unit of heat load, respectively. The number of days in different seasons.

6. The multi-stage rolling stochastic programming method for a regional integrated energy system according to claim 5, characterized in that: The cost of carbon reduction The reduction in carbon tax costs stems from the replacement of coal-fired boilers with centralized heating at energy hubs to provide heat load, and the reduction in carbon emissions due to the use of renewable energy. This can be expressed as: In the formula, , , , These represent the reduction in carbon emissions from heat, electricity, gas, and renewable energy sources in different seasons; , , These are the energy demands for heat, electricity, and gas on a typical day, where electricity and gas are primary energy inputs, and the secondary heat demand converted from gas and electricity is not included in the primary energy inputs. , , These represent the carbon emissions of heat, electricity, and gas on a typical day in different seasons. and These are the standard coal equivalent conversion factors for heat output and electrical output, respectively; and These are the CO2 emission coefficients for coal and natural gas, respectively. and These represent the carbon emissions from heat before and after the planning.

7. The multi-stage rolling stochastic programming method for a regional integrated energy system according to claim 3, characterized in that: The constraints on the energy equipment and energy pipelines are as follows: the constraints of the multi-energy flow model are multi-energy flow constraints; the capacity of the power pipelines corresponding to each pipeline type in stage s, and the capacity of the energy pipelines or energy equipment in the next stage should be greater than or equal to that in the previous stage; the upper limit of the flow of natural gas pipelines and heating pipelines should be greater than the flow calculation results of each pipeline to meet the load demand. The constraint condition at stage s is: In the formula, , , These are the upper limits for power flow in power lines, natural gas pipelines, and heating pipelines, respectively. Let K be the equipment capacity of substation k. The load provided by substation k; i is the i-th pipeline; The constraints for the next stage are: In the formula, and These represent the capacity of the energy pipelines and energy equipment during stage s; and These refer to the capacity of the energy pipelines and energy equipment in the S-1 stage; The restrictions on the type of energy equipment and investment cost are as follows: In the formula, The investment cost at stage s. For investment constraints; and These represent the wind power generation capacity of stage s and the previous stage of stage s, respectively.

Images