Optimization method and system for utilization of low-grade heat source by electrothermal synergy taking into account source-load uncertainty

Abstract

The present invention relates to a method and system for optimizing the utilization of low-grade heat sources through electric-heat collaboration, taking into account source-load uncertainty. The method comprises the following steps: randomly generating source and load output scenarios, and reducing the random scenarios; establishing an electric-heat collaborative planning model and constraints for the utilization of low-grade heat sources in a park, taking into account source-load uncertainty, and simplifying and solving the model to determine the energy hub equipment configuration capacity under different scenarios; obtaining the final energy hub equipment configuration by multiplying the energy hub equipment configuration capacity with the probability product of each scenario retained after scenario reduction; repeatedly calculating the final energy hub equipment configuration capacity until the error between two consecutive calculated equipment configuration results is less than a set value P_lim, and outputting the final final energy hub equipment configuration. The present invention considers source and load uncertainty, randomly generates source and load output scenarios based on predicted values and error distribution, reduces the random scenarios, and reduces computational complexity.

Description

Technical Field

[0001] The present invention belongs to the field of integrated energy system optimization, and in particular relates to a method and system for optimizing the utilization of low-grade heat sources in an electric-thermal coordinated manner taking into account source-load uncertainty. Background Art

[0002] The power system is currently undergoing a transformation from a traditional one to one that incorporates a high proportion of renewable energy. By 2050, the proportion of electricity in final energy consumption will rise from the current 26% to 45%-50%, a fundamental shift in the power system. The future power grid will feature both a high proportion of renewable energy and a high proportion of power electronics.

[0003] In 2018, the total length of centralized heating pipelines in Chinese cities reached 371,120 km, heating 878.085 million square meters. Centralized heating is the primary method of supply. Over the next 20 years, we must accelerate the electrification of heating, fully utilize renewable energy, and reduce carbon emissions from heating. my country boasts abundant low-grade heat sources, including geothermal, hydrothermal, low-temperature industrial waste heat, and a large amount of waste heat from data centers. Data indicates that during the winter heating season, approximately 4 billion gigajoules of low-grade industrial waste heat is generated in centralized heating areas in northern my country.

[0004] Currently, some low-grade heat sources, such as geothermal and industrial waste heat, have been used in district heating. However, these sources are typically supplied independently, which can lead to unstable heating parameters and hinder centralized scheduling. The electric-heat synergistic energy supply strategy is an effective way to utilize low-grade heat sources. By integrating these sources through energy hubs, it maximizes their utilization. Summary of the Invention

[0005] The purpose of this invention is to plan the equipment configuration of the energy hub station for the coordinated electric and thermal energy supply of the park, and to achieve low energy consumption and low carbon emissions in the park by maximizing the use of renewable energy and low-grade heat sources.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: a method for coordinating the utilization of low-grade heat sources in a park with electric and thermal energy taking into account the uncertainty of source and load, including using a scenario method to describe the uncertainty of electric load, renewable energy generation and low-grade heat sources; establishing an electric and thermal coordinating low-grade heat source utilization planning model; and the final equipment capacity result is the sum of the equipment configuration and probability product calculated for different scenarios.

[0007] Specifically:

[0008] A method for optimizing the utilization of low-grade heat sources by synergizing electricity and heat with consideration of source-load uncertainty, including

[0009] Randomly generate source and load output scenarios, and reduce random scenarios;

[0010] Establish an electric-heat collaborative planning model and constraints for the utilization of low-grade heat sources in the park, taking into account source-load uncertainty, simplify its solution, and determine the equipment configuration capacity of the energy hub station under different scenarios;

[0011] The final energy hub equipment configuration is obtained by multiplying the energy hub equipment configuration capacity and the probability of each scenario retained after scenario reduction.

[0012] Repeatedly calculate the final energy hub station equipment configuration capacity until the error of the equipment configuration results of two consecutive calculations is less than the set value , output the final energy hub equipment configuration for the last time.

[0013] In the above-mentioned optimization method for utilizing low-grade heat sources in electrothermal synergy taking into account source-load uncertainty, the scenario reverse reduction method RSA is used to perform random scenario reduction, specifically including:

[0014] Step 1.1: Use Gaussian distribution to describe the prediction errors of electricity load, renewable energy output, and low-grade heat source output:

[0015]

[0016] in: 、 、 are the prediction errors of electricity load, renewable energy output and low-grade heat source output, respectively. 、 、 represent the standard deviation of the prediction errors of electricity load, renewable energy output, and low-grade heat source output, respectively.

[0017] Step 1.2: Use the Monte Carlo method to generate error scenarios and add them to the predicted data to obtain a series of random scenario sets. .

[0018]

[0019]

[0020]

[0021]

[0022] in: represents the i-th random scene set at time t; 、 、 They represent the predicted output values of electric load, photovoltaic and low-grade heat source at time t respectively; 、 、 They represent the electric load value, photovoltaic and low-grade heat source output values under the i-th random scenario at time t respectively. is the total number of scenes.

[0023] Step 1.3, use the reverse reduction algorithm RSA to Random scenes are reduced to A certain scene.

[0024] In the above optimization method, when performing reverse reduction, it specifically includes

[0025] Step 1.3.1. Calculate the distance between each scene and other scenes and find the minimum value, which is recorded as ,,Scene and The distance between them is the paradigm distance, and each scene is generated with equal probability:

[0026]

[0027] in: Representation scene With scene distance, is the initial random scene probability.

[0028] Step 1.3.2: Calculate all scenarios The probability distance set of :

[0029]

[0030]

[0031] For the scene Minimum distance scene distance; For the scene The probability distance, is the set of all scenario probabilities.

[0032] Step 1.3.3, the scene Corresponding minimum distance scene Make reductions and add them to the reduced set In the scene The probability of correction is:

[0033]

[0034] in: To be reduced collection; For the corrected scene probability.

[0035] Steps 1.3.4, repeat the above process until the number of scenes is reduced to and obtain the occurrence probability of each scene.

[0036] In the above optimization method, the objective function of the electric-heat collaborative planning model is to minimize the comprehensive cost:

[0037]

[0038]

[0039]

[0040] in: is the comprehensive cost; is the annualized equipment cost, including investment cost and operation and maintenance costs ; are the gas purchase costs at time t, is the gas price at time t; is the cost of carbon emissions, is the carbon emission factor per unit of high power of natural gas, is the market carbon price.

[0041] In the above optimization method, the equipment cost model is determined based on the following formula:

[0042]

[0043]

[0044] in 、 、 、 They are the annualized construction investment costs per unit capacity of heat pumps, organic Rankine cycles, combined heat and power units, and gas boilers; 、 、 、 The rated capacity of heat pumps, organic Rankine cycles, combined heat and power units, and gas boilers; 、 、 、 is the operating power of the heat pump, organic Rankine cycle, cogeneration unit, and gas boiler at time t.

[0045] In the above optimization method, the constraints include equipment operation constraints, specifically:

[0046] Combined Heat and Power (CHP) unit operating constraints:

[0047]

[0048]

[0049]

[0050] in is the amount of natural gas input; Output electrical power to the CHP unit; is the power generation efficiency of the CHP unit, is the calorific value of natural gas; Waste heat for CHP unit power generation; is the heat loss rate; Output thermal power for CHP unit; is the heating coefficient.

[0051] Gas boiler (GB) operation constraints:

[0052]

[0053] In the formula is the thermal power output of GB; Input natural gas power for CB; is the GB energy conversion efficiency.

[0054] The heat pump HP operation constraints consider the relationship between input power, temperature and output power, temperature:

[0055]

[0056]

[0057]

[0058] in: Output heat energy to HP; Input electrical energy for the heat pump; is the conversion factor; 、 = is the temperature of the heat medium at the input and output sides of the heat pump. If the input power value of the heat pump is large, the waste heat flow cannot meet the heat pump's demand for absorbing enough heat, and it is in a low-efficiency operation state. There is a maximum constraint on the input power of the heat pump unit. is the waste heat input.

[0059] Organic Rankine Cycle (ORC) operating constraints:

[0060]

[0061] Where: is the ORC power generation efficiency, The low-grade heat source power input to the organic Rankine cycle requires the low-grade heat source temperature to be greater than 70°C.

[0062] In the above optimization method, the constraints also include electrical and thermal flow constraints, specifically:

[0063] The power flow adopts the DistFlow model:

[0064]

[0065]

[0066]

[0067]

[0068] in: 、 is the active and reactive power of the line; 、 are the active and reactive power of the line with the terminal node j respectively; 、 are the active and reactive power of the node respectively; is the square of the line current; 、 is the line impedance. By relaxing it into a second-order cone model, it has been shown that the relaxed distflow model can approximate the nonlinear power flow equation without loss in radial power grids.

[0069] The node power balance equation of the distribution network is:

[0070]

[0071] The heat network flow includes hydraulic model and thermal model hydraulic, the node flow balance equation in the model is:

[0072] Hydraulic model:

[0073]

[0074]

[0075]

[0076] Thermal model:

[0077]

[0078]

[0079]

[0080] Where: is the heat network correlation matrix, is the pipeline flow, and is the node flow; is the loop incidence matrix, Pipeline head loss; of which: is the head loss coefficient; is the specific heat capacity of water, 、 is the supply and return water temperature of node i at time t; 、 、 are the temperature at the beginning and end of pipe section i and the ambient temperature respectively;

[0081] In the above optimization method, when solving the electric-heat collaborative planning model, the first step is to perform piecewise linearization.

[0082] The piecewise linearization method for solving the heat pump model, such as Figure 3 As shown,

[0083] First, the equation is divided into k linear segments according to its domain of definition. k represents the number of linear segments. Generally speaking, the larger the value of k and the more segments there are, the higher the linear fitting accuracy will be.

[0084] Within each segment, the functional relationship is further linearly fitted, while ensuring that each function segment is closely connected.

[0085] The specific piecewise linearization process is as follows:

[0086] (1) Introducing variables , is a nonlinear single variable function.

[0087] (2) Approximate the nonlinear terms in the equation by piecewise linearization, determine the appropriate number of linearization segments, and set the horizontal coordinate interval [0, ] is divided into 1 sections.

[0088] (3) In Calculate the value of each segment point in the feasible region and obtain the corresponding vertical coordinate value.

[0089] (4) Introduce auxiliary 0-1 variables required for segmentation , which is used to indicate whether the point being sought falls within the linear segmented interval. If so, then ,otherwise .Will It is expressed by the following formula.

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096] in 、 is the horizontal coordinate of the beginning and end of the jth segment, 、 are the slope and intercept of the jth segment respectively.

[0097]

[0098] The heat pump COP exhibits a nonlinear relationship with both input and output temperatures. This study considers a steady-state model, assuming a constant input temperature at the heat source node and a constant output temperature at the load node. Based on the differences between the source and load nodes, the equation becomes a single-variable equation for the heat pump COP versus input or output temperature. Each node equation is then piecewise linearized using the aforementioned method.

[0099] In the above optimization method, an alternating iterative algorithm is used to process a certain segment after the piecewise linearization process, and the nonlinear , the alternating iterative method is used to solve it, specifically:

[0100] (1) Multiply the variables and linearize them

[0101]

[0102] 、 is the k-th iteration value, 、 Solve for the variables for the kth time.

[0103] Setting tolerances

[0104]

[0105] Repeat the iterative process until the error is less than the allowable value or the number of iterations reaches the maximum, the calculation ends, and the output is 、 value.

[0106] Iterative value update

[0107]

[0108] The iterative process uses the square root of the last iteration result and the iterative substitution value to update and increase the convergence speed.

[0109] A system comprising

[0110] The first module is configured to randomly generate source and load scenarios, and reduce the random scenarios;

[0111] The second module is configured to establish an electric-heat collaborative planning model and constraints for the utilization of low-grade heat sources in the park, taking into account source-load uncertainty, and simplify its solution to determine the equipment configuration capacity of the energy hub station under different scenarios;

[0112] The third module is configured to obtain the final energy hub station equipment configuration by multiplying the energy hub station equipment configuration capacity and the probability of each scenario remaining after scenario reduction;

[0113] The fourth module is configured to repeatedly calculate the final energy hub station equipment configuration capacity until the error of the equipment configuration results of two consecutive calculations is less than the set value , output the final energy hub equipment configuration for the last time.

[0114] Therefore, the present invention has the following advantages: 1. Taking into account the uncertainty of sources and loads, source and load output scenarios are randomly generated based on predicted values and error distributions, reducing random scenarios and computational complexity. 2. A coordinated electric-heat planning model is established that considers the utilization of low-grade heat sources. This model considers resource endowments, network trends, and the spatiotemporal correlations of energy supply components to optimize equipment configuration capacity. 3. The equipment configuration result is obtained by multiplying the optimization results under each random scenario with the scenario probability. The scenario is then regenerated and optimized until the equipment configuration result meets the error requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0115] Figure 1 This is a schematic diagram of the electric and thermal collaborative energy supply for the park of the present invention;

[0116] Figure 2 is an example of piecewise linearization of a nonlinear function of the present invention;

[0117] Figure 3 It is a flow chart of the alternating iterative algorithm of the present invention;

[0118] Figure 4 It is a diagram of an embodiment of the present invention;

[0119] Figure 5 (a) is a photovoltaic scene generation diagram according to an embodiment of the present invention;

[0120] FIG5( b ) is a photovoltaic scene reduction diagram according to an embodiment of the present invention;

[0121] FIG6 (a) is a diagram of the power supply equipment output in a maximum norm scenario according to an embodiment of the present invention;

[0122] FIG6( b ) is a diagram of the heating equipment output in a maximum norm scenario according to an embodiment of the present invention. DETAILED DESCRIPTION

[0123] The technical solution of the present invention will be further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0124] Example:

[0125] First, the principle of this method is introduced. The energy flow of the park's electric and thermal collaborative energy supply is as follows: Figure 1 The specific steps include:

[0126] Step 1: Use the Monte Carlo method to randomly generate source and load scenarios, and use the reverse scenario reduction method (RSA) to reduce the random scenarios;

[0127] Step 2: Considering the low-grade heat source resource endowment, equipment output, and power grid flow constraints, establish an electric-heat collaborative planning model for the park's low-grade heat source utilization that takes into account source-load uncertainty;

[0128] Step 3: Use piecewise linearization and alternating iterative algorithm to simplify and solve the nonlinear model;

[0129] Step 4: Optimize the energy hub station equipment configuration capacity under different scenarios, and calculate the equipment configuration capacity and the probability product to obtain the final energy hub station equipment configuration; repeat steps 1-4 until the error of the two equipment configuration results is less than the set value .

[0130] Step 5: Use the Monte Carlo method to randomly generate source and load output scenarios, define the maximum norm scenario as the extreme scenario, and verify the system operation stability under the extreme scenario.

[0131] The implementation of step 1 includes the following steps:

[0132] Step 1.1: Use Gaussian distribution to describe the prediction errors of electricity load, renewable energy output, and low-grade heat source output:

[0133]

[0134] in: 、 、 are the prediction errors of electricity load, renewable energy output and low-grade heat source output, respectively. 、 、 represent the standard deviation of the prediction errors of electricity load, renewable energy output, and low-grade heat source output, respectively.

[0135] Step 1.2: Use Monte Carlo method to randomly generate the load, renewable energy generation and low-grade heat source. Output scene.

[0136]

[0137]

[0138]

[0139]

[0140] in: represents the i-th random scene set at time t; 、 、 They represent the predicted output values of electric load, photovoltaic and low-grade heat source at time t respectively; 、 、 They represent the electric load value, photovoltaic and low-grade heat source output values under the i-th random scenario at time t respectively. is the total number of scenes.

[0141] Step 1.3.1. Calculate the distance between each scene and other scenes and find the minimum value, which is recorded as ,,Scene and The distance between them is the paradigm distance, and each scene is generated with equal probability:

[0142]

[0143] in: Representation scene With scene distance, is the initial random scene probability.

[0144] Step 1.3.2: Calculate all scenarios The probability distance set of :

[0145]

[0146]

[0147] in: For the scene Minimum distance scene distance; For the scene The probability distance, is the set of all scenario probabilities.

[0148] Step 1.3.3, the scene Corresponding minimum distance scene Make reductions and add them to the reduced set In the scene The probability of correction is:

[0149]

[0150] in: To be reduced collection; For the corrected scene probability.

[0151] Steps 1.3.4, repeat the above process until the number of scenes is reduced to and obtain the occurrence probability of each scene.

[0152] The implementation of step 2 includes the following steps:

[0153] Step 2.1: Establish an electric-heat collaborative planning model

[0154] Step 2.1.1. Set the objective function. The objective function is to minimize the comprehensive cost:

[0155]

[0156]

[0157]

[0158] in: is the comprehensive cost; is the annualized equipment cost, including investment cost and operation and maintenance costs ; are the gas purchase costs at time t, is the gas price at time t; is the cost of carbon emissions, is the carbon emission factor per unit of high power of natural gas, is the market carbon price.

[0159] Equipment cost model:

[0160]

[0161]

[0162] in 、 、 、 They are the annualized construction investment costs per unit capacity of heat pumps, organic Rankine cycles, combined heat and power units, and gas boilers; 、 、 、 The rated capacity of heat pumps, organic Rankine cycles, combined heat and power units, and gas boilers; 、 、 、 is the operating power of the heat pump, organic Rankine cycle, cogeneration unit, and gas boiler at time t.

[0163] Step 2.2.2: Equipment operation constraints

[0164] (1) Combined heat and power units

[0165] Combined Heat and Power (CHP) unit operating constraints:

[0166]

[0167]

[0168]

[0169] in is the amount of natural gas input; Output electrical power to the CHP unit; is the power generation efficiency of the CHP unit, is the calorific value of natural gas; Waste heat for CHP unit power generation; is the heat loss rate; Output thermal power for CHP unit; is the heating coefficient.

[0170] (2) Gas boiler

[0171] Gas boiler (GB) operation constraints:

[0172]

[0173] In the formula is the thermal power output of GB; Input natural gas power for CB; is the GB energy conversion efficiency.

[0174] (3) Heat pump

[0175] Heat pump (HP) operating constraints consider the relationship between input power, temperature and output power, temperature:

[0176]

[0177]

[0178]

[0179] in: Output heat energy to HP; Input electrical energy for the heat pump; is the conversion factor; 、 = is the temperature of the heat medium at the input and output sides of the heat pump. If the input power value of the heat pump is large, the waste heat flow cannot meet the heat pump's demand for absorbing enough heat, and it is in a low-efficiency operation state. There is a maximum constraint on the input power of the heat pump unit. is the waste heat input.

[0180] (4) Organic Rankine cycle

[0181] Organic Rankine Cycle (ORC) operating constraints:

[0182]

[0183] Where: is the ORC power generation efficiency, The low-grade heat source power input to the organic Rankine cycle requires the low-grade heat source temperature to be greater than 70°C.

[0184] Step 2.2.4: Electrical and thermal flow constraints

[0185] The power flow adopts the DistFlow model:

[0186]

[0187]

[0188]

[0189]

[0190] in: 、 is the active and reactive power of the line; 、 are the active and reactive power of the line with the terminal node j respectively; 、 are the active and reactive power of the node respectively; is the square of the line current; 、 is the line impedance. By relaxing it into a second-order cone model, it has been shown that the relaxed distflow model can approximate the nonlinear power flow equation without loss in radial power grids.

[0191] The node power balance equation of the distribution network is:

[0192]

[0193] The heat network flow includes hydraulic model and thermal model hydraulic, the node flow balance equation in the model is:

[0194]

[0195] in: is the heat network correlation matrix, is the pipeline flow rate, is the node traffic.

[0196] Circuit pressure balance equation:

[0197]

[0198] in: is the loop incidence matrix, Pipe head loss.

[0199] Head loss equation:

[0200]

[0201] in: is the head loss coefficient.

[0202] Relationship between node power, temperature and flow in thermal model:

[0203]

[0204] is the specific heat capacity of water, 、 is the supply and return water temperature of node i at time t.

[0205] Pipeline temperature drops:

[0206]

[0207] 、 、 are the temperature at the beginning and end of pipe section i and the ambient temperature respectively;

[0208] Thermal topology constraints:

[0209]

[0210] Step 3 includes the following specific steps:

[0211] Step 3.1 Piecewise linearization

[0212] For the heat pump model, piecewise linearization is performed, such as Figure 3 As shown, first divide the equation into k linear segments according to its domain, where k represents the number of linear segments. Generally speaking, the larger the value of k and the more segments there are, the higher the linear fitting accuracy. Within each segment, further linear fitting is performed on the function relationship, while ensuring that each segment of the function is closely connected. The specific piecewise linearization process is as follows: Figure 2 The specific process is as follows:

[0213] (1) Introducing variables , is a nonlinear single variable function.

[0214] (2) Approximate the nonlinear terms in the equation by piecewise linearization, determine the appropriate number of linearization segments, and set the horizontal coordinate interval [0, ] is divided into 1 sections.

[0215] (3) In Calculate the value of each segment point in the feasible region and obtain the corresponding vertical coordinate value.

[0216] (4) Introduce auxiliary 0-1 variables required for segmentation , which is used to indicate whether the point being sought falls within the linear segmented interval. If so, then ,otherwise .Will It is expressed by the following formula.

[0217]

[0218]

[0219]

[0220]

[0221]

[0222]

[0223] in 、 is the horizontal coordinate of the beginning and end of the jth segment, 、 are the slope and intercept of the jth segment respectively.

[0224]

[0225] The heat pump COP exhibits a nonlinear relationship with both input and output temperatures. This study considers a steady-state model, assuming a constant input temperature at the heat source node and a constant output temperature at the load node. Based on the differences between the source and load nodes, the equation becomes a single-variable equation for the heat pump COP versus input or output temperature. Each node equation is then piecewise linearized using the aforementioned method.

[0226] Step 3.2 Alternating Iteration Algorithm

[0227]

[0228] The above equation multiplies the heat pump input power and the heat pump COP, which is nonlinear. According to game theory, this model has a unique equilibrium point, so the alternating iterative method can be used to solve it. The process is as follows: Figure 3 The specific solution steps are as follows:

[0229] (1) Multiply the variables and linearize them

[0230]

[0231] 、 is the k-th iteration value, 、 Solve for the variables for the kth time.

[0232] Setting tolerances

[0233]

[0234] Repeat the iterative process until the error is less than the allowable value or the number of iterations reaches the maximum, the calculation ends, and the output is 、 value.

[0235] Iterative value update

[0236]

[0237] The iterative process uses the square root of the last iteration result and the iterative substitution value to update and increase the convergence speed.

[0238] The implementation of step 4 includes the following steps:

[0239] Step 4.1: Optimize the energy hub equipment configuration capacity under different scenarios, and sum the calculated equipment configuration capacity and the probability product to obtain the final energy hub equipment configuration.

[0240] Step 4.2: Repeat steps 1-4 until the error between the two device configuration results is less than the set value. .

[0241] The implementation of step 5 includes the following steps:

[0242] Step 5.1: Use the Monte Carlo method to randomly generate source and load output scenarios and find the maximum norm scenario with the largest distance from the source and load prediction scenario.

[0243] Step 5.2: Under the existing equipment configuration, with minimizing the park operating cost as the objective function, optimize the equipment output and verify the system operation stability under extreme scenarios.

[0244] This embodiment also relates to a system, including

[0245] The first module is configured to randomly generate source and load scenarios, and reduce the random scenarios;

[0246] The second module is configured to establish an electric-heat collaborative planning model and constraints for the utilization of low-grade heat sources in the park, taking into account source-load uncertainty, and simplify its solution to determine the equipment configuration capacity of the energy hub station under different scenarios;

[0247] The third module is configured to obtain the final energy hub station equipment configuration by multiplying the energy hub station equipment configuration capacity and the probability of each scenario remaining after scenario reduction;

[0248] The fourth module is configured to repeatedly calculate the final energy hub station equipment configuration capacity until the error of the equipment configuration results of two consecutive calculations is less than the set value , output the final energy hub station equipment configuration for the last time. 2. The following is a description of a specific case using the present invention.

[0249] The embodiment is set as an 8-node power system and a 9-node thermal system, such as Figure 4 As shown, nodes 7 and 8 in the power system are photovoltaic and wind power nodes respectively, and the rest are electric load nodes. Nodes VII, VIII, and IX in the thermal system are industrial waste heat, data center waste heat, and geothermal heat respectively, and the rest are thermal load nodes.

[0250] 3. Planning results:

[0251] 3.1 Scenario Generation and Reduction Results

[0252] The random scenarios of electric load, wind power, and low-grade heat sources generated by the Monte Carlo method, and the generation and reduction of photovoltaic scenarios are shown in Figures 5 (a) and (b).

[0253] 3.2 Device Configuration Results and Device Operation Strategy for Maximum Norm Scenario

[0254] Table 1 shows the equipment configuration results for the energy hub station, which utilizes low-grade heat sources in a coordinated electric-thermal planning approach, taking into account both source and load uncertainties. Figures 6(a) and 6(b) show the equipment output under the maximum norm scenario. This equipment configuration meets the load demand, with typical daily operating costs of 1.08 × 10⁵ and 1.32 × 10⁵ yuan under the predicted and norm scenarios, respectively.

[0255] Table 1 Equipment capacity configuration

[0256]

[0257] The specific embodiments described herein are merely illustrative of the spirit of the present invention. Persons skilled in the art may make various modifications, additions, or substitutions to the described specific embodiments without departing from the spirit of the present invention or exceeding the scope of the appended claims.

Claims

1. A method for optimizing the utilization of low-grade heat sources by combining electric and thermal energy with consideration of source-load uncertainty, characterized in that: include Randomly generate source-load output scenarios and use the scenario reverse reduction method (RSA) to perform random scenario reduction. Establish an electric-heat collaborative planning model and constraints for the utilization of low-grade heat sources in the park, taking into account source-load uncertainty, simplify the solution, and determine the equipment configuration capacity of the energy hub station under different scenarios; The final energy hub equipment configuration is obtained by multiplying the energy hub equipment configuration capacity and the probability of each scenario retained after scenario reduction. Repeatedly calculate the final energy hub station equipment configuration capacity until the error of the equipment configuration results of two consecutive calculations is less than the set value , output the final energy hub equipment configuration for the last time; The scene reverse reduction method RSA is used to perform random scene reduction, specifically including: Step 1.1: Use Gaussian distribution to describe the prediction errors of electricity load, renewable energy output, and low-grade heat source output: in: 、 、 are the prediction errors of electricity load, renewable energy output and low-grade heat source output, respectively. 、 、 denote the standard deviation of the prediction errors of electric load, renewable energy output, and low-grade heat source output, respectively; Step 1.2: Use the Monte Carlo method to generate error scenarios and add them to the predicted data to obtain a series of random scenario sets. ; in: represents the i-th random scene set at time t; 、 、 They represent the predicted output values of electric load, photovoltaic and low-grade heat source at time t respectively; 、 、 They represent the electric load value, photovoltaic and low-grade heat source output values under the i-th random scenario at time t respectively; is the total number of scenes; Step 1.3, use the reverse reduction algorithm RSA to Random scenes are reduced to A certain scenario; The objective function of the electricity-heat collaborative planning model is to minimize the comprehensive cost: in: is the comprehensive cost; is the annualized equipment cost, including investment cost and operation and maintenance costs ; are the gas purchase costs at time t, is the gas price at time t; is the cost of carbon emissions, is the carbon emission factor per unit of high power of natural gas, is the market carbon price.

2. The method for optimizing utilization of low-grade heat sources by combining electricity and heat with source-load uncertainty according to claim 1 is characterized in that: When performing scene generation and reduction, it specifically includes Step 1.3.

1. Calculate the distance between each scene and other scenes and find the minimum value, which is recorded as , scene and The distance between them is the paradigm distance, and each scene is generated with equal probability: in: Representation scene With scene distance, is the initial random scene probability; Step 1.3.2: Calculate all scenarios The probability distance set of : in: For the scene Minimum distance scene distance; For the scene The probability distance, is the set of all scenario probabilities; Step 1.3.3, the scene Corresponding minimum distance scene Make reductions and add them to the reduced set In the scene The probability of correction is: in: To be reduced collection; For the corrected scene probability; Steps 1.3.4, repeat the above process until the number of scenes is reduced to and obtain the occurrence probability of each scene.

3. The method for optimizing the utilization of low-grade heat sources by combining electricity and heat with source-load uncertainty according to claim 1 is characterized in that: The equipment cost model is determined based on the following formula: in 、 、 、 They are the annualized construction investment costs per unit capacity of heat pumps, organic Rankine cycles, combined heat and power units, and gas boilers; 、 、 、 The rated capacity of heat pumps, organic Rankine cycles, combined heat and power units, and gas boilers; 、 、 、 is the operating power of the heat pump, organic Rankine cycle, cogeneration unit, and gas boiler at time t.

4. The method for optimizing the utilization of low-grade heat sources by combining electricity and heat with source-load uncertainty according to claim 1 is characterized in that: Constraints include equipment operation constraints, specifically: Combined Heat and Power (CHP) unit operating constraints: in is the amount of natural gas input; Output electrical power to the CHP unit; is the power generation efficiency of the CHP unit, is the calorific value of natural gas; Waste heat for CHP unit power generation; is the heat loss rate; Output thermal power for CHP unit; is the heating coefficient; Gas boiler (GB) operation constraints: In the formula is the thermal power output of GB; Input natural gas power for CB; is the GB energy conversion efficiency; Heat pump (HP) operating constraints consider the relationship between input power, temperature and output power, temperature: in: Output heat energy to HP; Input electrical energy for the heat pump; is the conversion factor; 、 is the temperature of the heat medium at the input and output sides of the heat pump; if the heat pump input power value is large, the waste heat flow cannot meet the heat pump's demand for absorbing enough heat, and it is in a low-efficiency operation state; the heat pump unit has a maximum constraint on the input power, is the waste heat input; Organic Rankine cycle (ORC) output constraints: Where: is the ORC power generation efficiency, Low-grade heat source power input to the organic Rankine cycle.

5. The method for optimizing utilization of low-grade heat sources by combining electricity and heat with consideration of source-load uncertainty according to claim 1 is characterized in that: Constraints also include electrical and thermal flow constraints, specifically: The power flow adopts the DistFlow model: in: 、 is the active and reactive power of the line; 、 are the active and reactive power of the line with the terminal node j respectively; 、 are the active and reactive power of the node respectively; is the square of the line current; 、 is the line impedance; it is transformed into a second-order cone model through relaxation. It has been proven that in radial power grids, the relaxed distflow model can losslessly approximate the nonlinear power flow equation; The node power balance equation of the distribution network is: The heat network flow includes hydraulic model and thermal model hydraulic, the node flow balance equation in the model is: Hydraulic model: Thermal model: Where: is the heat network correlation matrix, is the pipeline flow, and is the node flow; is the loop incidence matrix, Pipeline head loss; of which: is the head loss coefficient; is the specific heat capacity of water, 、 is the supply and return water temperature of node i at time t; 、 、 are the starting and ending temperatures of pipe section i and the ambient temperature, respectively.

6. The method for optimizing utilization of low-grade heat sources by combining electricity and heat with source-load uncertainty according to claim 1 is characterized in that: When solving the electric-heat collaborative planning model, piecewise linearization is first performed. For the nonlinear part of the heat pump nonlinear model: The heat pump COP in the above equation has a nonlinear relationship with the input and output temperatures. Considering the steady-state model, the input temperature of the heat pump at the heat source node and the output temperature of the heat pump at the load node are set to be constant. Based on the difference between the source and load nodes, the equation becomes a single variable equation of the heat pump COP and the input or output temperature. The equations of each node are solved by piecewise linearization according to the above method.

7. The method for optimizing utilization of low-grade heat sources by combining electricity and heat with consideration of source-load uncertainty according to claim 1 is characterized in that: After the piecewise linearization processing, the alternating iterative algorithm is used to process it, and the nonlinear , the alternating iterative method is used to solve it, specifically: (1) Multiply the variables and linearize them 、 is the k-th iteration value, 、 Solve for the variables for the kth time; Setting tolerances Repeat the iterative process until the error is less than the allowable value or the number of iterations reaches the maximum, the calculation ends, and the output is 、 value; Iterative value update The iterative process uses the square root of the last iteration result and the iterative substitution value to update and increase the convergence speed.

8. A system, characterized in that: include The first module is configured to randomly generate source and load scenarios, and reduce the random scenarios; The second module is configured to establish an electric-heat collaborative planning model and constraints for the utilization of low-grade heat sources in the park, taking into account source-load uncertainty, and simplify the solution to determine the configuration capacity of energy hub equipment under different scenarios; The third module is configured to obtain the final energy hub station equipment configuration by multiplying the energy hub station equipment configuration capacity and the probability of each scenario remaining after scenario reduction; The fourth module is configured to repeatedly calculate the final energy hub station equipment configuration capacity until the error of the equipment configuration results of two consecutive calculations is less than the set value , output the final energy hub equipment configuration for the last time; The scene reverse reduction method RSA is used to perform random scene reduction, specifically including: Step 1.1: Use Gaussian distribution to describe the prediction errors of electricity load, renewable energy output, and low-grade heat source output: in: 、 、 are the prediction errors of electricity load, renewable energy output and low-grade heat source output, respectively. 、 、 denote the standard deviation of the prediction errors of electric load, renewable energy output, and low-grade heat source output, respectively; Step 1.2: Use the Monte Carlo method to generate error scenarios and add them to the predicted data to obtain a series of random scenario sets. ; in: represents the i-th random scene set at time t; 、 、 They represent the predicted output values of electric load, photovoltaic and low-grade heat source at time t respectively; 、 、 They represent the electric load value, photovoltaic and low-grade heat source output values under the i-th random scenario at time t respectively; is the total number of scenes; Step 1.3, use the reverse reduction algorithm RSA to Random scenes are reduced to A certain scenario; The objective function of the electricity-heat collaborative planning model is to minimize the comprehensive cost: in: is the comprehensive cost; is the annualized equipment cost, including investment cost and operation and maintenance costs ; are the gas purchase costs at time t, is the gas price at time t; is the cost of carbon emissions, is the carbon emission factor per unit of high power of natural gas, is the market carbon price.

Images