Multi-energy coupling medium-deep layer buried pipe group and power grid cooperative regulation and control method and system

Abstract

The invention discloses a multi-energy coupling middle-deep layer buried pipe group and power grid cooperative regulation and control method and system, and the method comprises the steps: designing a hierarchical regulation and control architecture of bottom layer single pipe real-time control-middle layer pipe group heat balance-top layer system planning, and fusing model prediction control MPC and a deep reinforcement learning algorithm to carry out multi-time scale regulation and control; the method comprises the following steps: constructing a geothermal-wind-solar-energy storage complementary collaborative model, analyzing an energy conversion and distribution mechanism under multiple time scales based on a system dynamics and energy quality cascade matching theory, and performing effective collaboration among different energy forms by establishing a thermal-electric-energy storage dynamic balance strategy. A carbon transaction cost prediction module and an electric power spot price prediction module are integrated; according to the method, motion space dimension reduction is carried out by combining a physical constraint embedding strategy based on deep reinforcement learning, solution tool integration is carried out by combining a reinforcement learning optimization method and an efficient solver, a multi-disturbance scene verification platform is constructed, complex working conditions of power grid fluctuation and permeability mutation are simulated, robustness testing is carried out, and a target model is output.

Description

Technical Field

[0001] The present invention relates to the field of coordinated regulation, and in particular to a method and system for coordinated regulation of a multi-energy coupled medium-deep buried pipe group and a power grid. Background Art

[0002] In recent years, significant progress has been made in operational strategy research regarding single-tube thermal storage control and parameter optimization. However, existing research indicates that relying solely on start-stop regulation results in an average annual energy loss of 12%-15%. This suggests that the operational strategy for mid- to deep-layer heat extraction is crucial to system performance.

[0003] Existing control strategies face three bottlenecks: First, multi-timescale response mismatches and a lack of a cross-level physical-information bidirectional coupling mechanism lead to poor information flow between multi-layer control systems, hindering the coordination of global energy allocation. Second, the understanding of the energy gradient coordination mechanism between geothermal energy and wind and solar energy is insufficient. Carbon trading and price fluctuations in the electricity spot market further increase the complexity of dispatch, and an optimization solution that balances economy, low carbon emissions, and stability has not yet been established. Third, the solvability barrier of high-dimensional dynamic optimization decision space makes conventional algorithms difficult to meet real-time control requirements under the constraints of time-varying geological parameters, complex pipeline network topologies, and multiple objectives. It is necessary to establish an optimization and solution method that combines intelligent control theory with multi-objective coordination to achieve adaptive control of pipeline groups and promote efficient coordination and stable interaction between diverse renewable energy sources and the power grid.

[0004] For the operation of medium- and deep-layer closed underground pipe systems and multi-energy systems, relevant evaluation indicators such as annual sustainability, seasonal non-guarantee rate, and lifecycle cost have been proposed. Furthermore, optimization configuration methods based on source-side temperature and flow have been developed, as well as multi-energy system hierarchical control strategies that consider multiple objectives such as renewable energy consumption, grid interaction, and economic efficiency. However, existing strategies mostly focus on the single-pipe scale, and pipe group-level control faces the multi-time scale mismatch between minute-level pipeline network fluctuations and cross-seasonal heat storage cycles. In terms of energy system coordination, preliminary results have been achieved in the study of the intraday complementarity between medium- and deep-layer geothermal energy and photovoltaic and shallow geothermal energy. Existing coupled energy storage system design methods have raised the clean power consumption rate to new heights. However, there is still a theoretical research gap in the dynamic interaction mechanism between medium- and deep-layer pipe groups and the power grid, and it is urgent to break through the optimization barriers of multi-level gradient coordination and the dual constraints of market and physics. Summary of the Invention

[0005] The purpose of the present invention is to provide a multi-energy coupled method for coordinated control of a medium-deep buried pipe group and a power grid.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] The present invention comprises the following steps:

[0008] A hierarchical control architecture consisting of "bottom-level single-pipe real-time control - middle-level pipe group heat balance - top-level system planning" was designed, integrating model predictive control (MPC) with deep reinforcement learning algorithms for multi-time-scale control. The hierarchical control architecture consists of bottom, middle, and top layers. The bottom layer uses model predictive control to optimize the distribution of energy stations and satellite stations based on heat load prediction, achieving real-time and precise control of single pipes. The middle layer coordinates heat distribution between pipe groups through a heat balance algorithm. The top layer uses system planning methods for overall energy management and optimization, establishing a rolling time-domain optimization feedback mechanism.

[0009] Construct a complementary synergistic model for geothermal, wind, solar, and energy storage. Based on system dynamics and energy quality ladder matching theory, analyze the energy conversion and distribution mechanism under multiple time scales. By establishing a dynamic balance strategy for heat, electricity, and storage, effectively coordinate different energy forms, and integrate carbon trading cost prediction and electricity spot price prediction modules.

[0010] A deep reinforcement learning-based approach combined with a physical constraint embedding strategy is used to reduce the action space dimension, obtain efficiency and physical feasibility, and integrate solution tools using reinforcement learning optimization methods and efficient solvers. A multi-disturbance scenario verification platform is constructed to simulate complex operating conditions such as grid fluctuations and sudden changes in penetration rate, conduct robustness tests, and output the target model.

[0011] Furthermore, the method for designing the bottom layer includes:

[0012] A system dispatch control method based on model prediction is adopted. The heat load demand of each heating station in the area is first predicted, and then optimization analysis and decision-making are carried out based on the prediction. The dispatch control method is categorized according to the time dimension and the space dimension. The space dimension mainly includes the equipment receiving the control command, ground source heat pumps, satellite stations, and energy stations. In the time dimension, it is divided into optimization control under short time scale and dispatch optimization under medium and long time scale according to the optimization analysis scale.

[0013] The top layer of the heating system first decomposes tasks based on energy production and orderly energy consumption plans. In the operation scheduling module, a control strategy is generated based on the specific conditions of multiple heat sources and the operating status of each device. In the intraday real-time optimization module, control targets are generated through real-time online monitoring of equipment and target optimization, and control actions are generated using advanced control methods. After completing the top-down scheduling and control process, the system collects monitoring data from lower-level equipment in real time and provides online feedback to the upper layer, realizing a closed-loop scheduling and control process. The upper layer links are corrected based on the feedback from the lower layers.

[0014] The heating system based on model prediction is controlled by optimized control logic, in which the optimized control logic is as follows: the future heat load at the thermal power station level is predicted through load forecasting technology, and the heating load distribution of the energy station and satellite station is determined in combination with the heating capacity and heating cost curve of the energy station and satellite station. Under the premise of maximizing the heating capacity of geothermal heat sources, efficient heating is coordinated with the heat sources of multiple energy stations to achieve low-carbon operation of the heating system; based on the heat load distribution, the scheduling platform gives the heating flow and supply and return water temperature of each energy station.

[0015] Furthermore, the method for designing the middle layer includes:

[0016] To regulate the heat distribution between different pipe groups, the host computer platform calculates the comprehensive physical property coefficient of the terminal room, classifies heat users, determines the regulation cycle and the target return water temperature based on historical heating parameter information;

[0017] When the heat demand of the heat user changes, the return water temperature is adjusted to adapt to the heat demand. The upper computer platform issues an adjustment instruction to change the opening of the intelligent valve to change the return water temperature of the heat user to reach the target return water temperature. When the real-time return water temperature meets the target and When the heat user's intelligent valve adjustment ends, the heat user's real-time return water temperature is The target value of the real-time return water temperature of the heat user is The return water temperature deviation threshold is ξ 2h , the actual temperature of the input water is The target temperature of the input water is The temperature deviation threshold is ξ in ;

[0018] On the contrary, continue to adjust the valve opening, ξ 2h Different systems may have different thresholds. The change in the opening of the heat user's intelligent valve will cause the pressure difference and impedance of the entire network to change. At this time, the water pump of the thermal power station will perform linked frequency conversion regulation to change the flow rate and restore the pressure difference and impedance of the network to the set value. When the network impedance and pressure difference meet the requirements, the system will automatically adjust the flow rate and return to the set value. |SS op |<σ s |, and the valve opening that is most unfavorable to the heat user is the maximum limit V z =V max When the frequency conversion is ended, otherwise the frequency conversion will continue; the actual pressure difference of the pipe network is ΔP, and the set value of the pipe network pressure difference is The deviation threshold of the pipe network pressure difference is σ P The actual impedance of the pipe network is S, and the set value of the pipe network impedance is S op , the deviation threshold of the pipe network impedance is σ s .

[0019] Furthermore, the method for designing the top layer includes:

[0020] The planning of a multi-energy coordinated integrated energy system is a conversion process between multiple energy sources. The system planning method abstracts an integrated energy system into a dual-port network of energy input and output, where multiple energy sources are converted, distributed, and stored within the integrated energy system.

[0021] Connect the input end of the system planning method to the energy network, input the corresponding electricity, gas, and oil energy at the input end, and output the energy in the form of electricity, heat, and cooling at the output end; based on the actual situation of the operation stage, establish an optimization planning model of the integrated energy system "planning and operation integration";

[0022] The optimization planning involves the optimal configuration scheme of equipment capacity and quantity, as well as the optimization of equipment cooling, heating and electricity operation mode. According to the decomposition and coordination idea, the optimization planning is transformed into a two-level planning model; the upper-level planning takes the maximization of the net present value within the integrated energy system as the goal, and carries out the optimal configuration of equipment capacity and quantity; the lower-level optimization corresponds to the optimization of the integrated energy system operation strategy. Based on the equipment capacity and quantity optimized in the upper-level, the minimum operating cost of the integrated energy system and the maximum proportion of electricity in the total energy consumption are objectively weighted to form the objective function. According to the equipment operation constraints, the cooling, heating and electricity operation mode of the equipment is optimized, and the optimization target is passed to the upper-level optimization, and the internal rate of return is calculated. Through the iterative optimization of the upper and lower levels, the optimal "planning and operation integration" integrated energy system optimization planning scheme is obtained.

[0023] Furthermore, the method of integrating the carbon transaction cost forecasting and electricity spot price forecasting modules includes:

[0024] On the TRNSYS platform, geothermal, wind, solar, and energy storage are coupled. The multi-energy coupled heating and cooling system model includes an energy consumption module, a cost module, and an output module. The energy consumption module is used to calculate the annual energy consumption of each cooling and heating source subsystem. The cost module uses time-of-use electricity and gas prices to calculate the annual operating costs. The output module outputs the capacity ratio, initial investment, operating energy consumption, cooling / heating, total initial investment, annual operating costs, annual carbon emissions, and life cycle costs of each subsystem.

[0025] In the multi-energy coupled heating and cooling system model, the GenOpt optimization tool connected to the TrnOpt module in TRNSYS is used to optimize variables.

[0026] Incorporating carbon emissions into the geothermal-wind-solar-energy storage system, a carbon emission cost model is established, expressed as:

[0027]

[0028] Among them, the basic carbon emission quota is Q0, the equivalent actual carbon emission is Q1, and the carbon emission cost is F co2 , the carbon emission cost coefficients are C1, C2, C3, C4 respectively, the primary boundary quota of carbon emission is, and the secondary boundary quota of carbon emission is Q 02 ;

[0029] When the emission Q1 is lower than Q0, the geothermal-wind-energy storage system sells the excess emission quota; when the emission Q1 is higher than Q0, the geothermal-wind-energy storage system needs to purchase carbon emission quotas, and as the purchased quota increases, the unit purchase cost increases; when Q1 < Q0, it means that when the emission does not exceed the basic quota Q0, the surplus quota is sold; when the emission exceeds different quotas, the excess part purchases carbon emission quotas according to the corresponding emission cost;

[0030] The electricity spot market conducts real-time trading of electricity in a short time cycle. The electricity spot price reflects the dynamic balance of electricity supply and demand and is affected by multiple factors, including weather conditions, load demand, fuel prices, generation capacity, network constraints, and market participants;

[0031] According to the advantages of different methods, a hybrid model of time series models and machine learning models, multi-model integration, and decomposition-based methods is proposed. The hybrid model effectively captures the complex characteristics of electricity spot prices; among them, the hybrid of time series models and machine learning models: first, use the time series model to extract the linear characteristics of price data, and then use the machine learning model to learn the non-linear characteristics of the residuals; multi-model integration: perform weighted averaging on the prediction results of multiple different models; decomposition-based method: first decompose the original price data into components of different frequencies, then establish prediction models for each component, and finally combine the prediction results of each component;

[0032] By establishing a thermal-electricity-storage dynamic balance strategy, effective coordination between different energy forms is achieved, and the carbon trading cost prediction and electricity spot price prediction modules are integrated to ensure the optimal scheduling of the energy system under different market environments and operating conditions.

[0033] Furthermore, the method for dimension reduction of the action space includes:

[0034] Based on the general framework of a multi-objective evolutionary algorithm for online target dimensionality reduction, the ε-MOSFLA algorithm is combined with the SORA1 and SORA2 target dimensionality reduction algorithms to construct online target dimensionality reduction algorithms that can meet different requirements. The online target dimensionality reduction algorithms include the SO1-MOSFLA algorithm and the SO2-MOSFLA algorithm. The SO1-MOSFLA algorithm uses SORA1 to reduce a fixed number of targets each time until the desired number of targets is reached, and no targets are deleted. The SO2-MOSFLA algorithm uses SORA2 to adaptively find unimportant targets and delete them according to a given error threshold until no more targets can be deleted or the number of targets is reduced to two.

[0035] Based on the online target dimensionality reduction algorithm, a target integration strategy is proposed. The targets to be deleted are classified according to their importance indicators. The importance indicators are integrated into one target through weighted accumulation and added to the target subset. This strategy solves two types of redundant and non-redundant high-dimensional multi-objective optimization problems.

[0036] This paper proposes a method based on sparse feature selection to measure the importance of the target by utilizing the geometric structure characteristics of the approximate solution set and the Pareto dominance relationship, and constructs a target dimensionality reduction algorithm for solving two different requirements.

[0037] Based on the idea of ​​sparse feature selection, a target preference ranking evaluation algorithm is proposed. When the error of the original problem does not exceed the error threshold, the target preference ranking evaluation algorithm starts from the original target set and only reduces one target at a time. The comprehensive ranking of the targets in terms of importance in each dimensionality reduction process is statistically analyzed to determine the ranking of each target.

[0038] Furthermore, the robustness testing method includes:

[0039] The labels obtained by classifying the U-shaped medium-deep buried pipe underground heat exchange model are obtained, and a local substitution model with highly similar classification capabilities to the U-shaped medium-deep buried pipe underground heat exchange model is constructed. Adversarial samples are constructed based on the output of the substitution model, and the transferability of the adversarial samples is used to achieve robustness testing of the target model.

[0040] Secondly, the multi-energy coupled medium-deep buried pipeline group and power grid coordinated control system includes:

[0041] Hierarchical multi-scale control module: This module is used to design a hierarchical control architecture consisting of "bottom-level single-tube real-time control - middle-level tube group thermal balance - top-level system planning." It integrates model predictive control (MPC) and deep reinforcement learning algorithms for multi-time-scale control. The hierarchical control architecture includes bottom, middle, and top layers. The bottom layer uses model predictive control to optimize the distribution of energy stations and satellite stations based on heat load prediction, achieving real-time and precise control of single tubes. The middle layer coordinates heat distribution between tube groups through a heat balance algorithm. The top layer uses system planning methods for overall energy management and optimization, establishing a rolling time-domain optimization feedback mechanism.

[0042] Collaborative Forecasting Module: This module is used to build a complementary synergistic model for geothermal, wind, solar, and energy storage. Based on system dynamics and energy quality ladder matching theory, it analyzes energy conversion and distribution mechanisms at multiple time scales. By establishing a dynamic balance strategy for heat, electricity, and storage, it effectively coordinates different energy forms and integrates carbon trading cost prediction and electricity spot price prediction modules.

[0043] Dimensionality reduction testing module: It is used to reduce the dimension of the action space by using a deep reinforcement learning-based combined with physical constraint embedding strategy to obtain efficiency and physical feasibility. It combines reinforcement learning optimization methods and efficient solvers to integrate solution tools, build a multi-disturbance scenario verification platform, simulate complex working conditions such as power grid fluctuations and sudden changes in penetration rate, conduct robustness tests, and output the target model.

[0044] The beneficial effects of the present invention are:

[0045] The present invention is a method and system for coordinated control of a multi-energy coupled medium-deep buried pipe group and a power grid. Compared with the prior art, the present invention has the following technical effects:

[0046] The present invention can improve the accuracy of the construction of the underground heat exchange model of the U-shaped medium-deep buried pipe through hierarchical regulation, multi-time scale regulation, construction of complementary collaborative models, energy conversion and distribution mechanism, effective coordination, integrated prediction module, action space dimensionality reduction and robustness test steps, thereby improving the accuracy of the construction of the underground heat exchange model of the U-shaped medium-deep buried pipe, optimizing the construction of the underground heat exchange model of the U-shaped medium-deep buried pipe, greatly saving resources, improving work efficiency, and realizing the scientific construction of the underground heat exchange model of the U-shaped medium-deep buried pipe, and real-time monitoring of the underground heat exchange model of the U-shaped medium-deep buried pipe. The construction of an underground heat exchange model of buried pipes is used to carry out hierarchical control and integrated prediction modules; by constructing a three-level intelligent control architecture of "single pipe-pipe group-system", integrating multi-dimensional control theory and efficient optimization algorithm, the problem of real-time solution of high-dimensional decision-making is overcome, and minute-level thermal balance response is achieved; at the same time, based on the complementary characteristics of geothermal, wind, solar and energy storage, a multi-time scale collaborative model covering day-ahead, intraday and real-time is established, which greatly improves the overall energy efficiency of the system and effectively optimizes the peak-valley regulation capability of the power grid, providing cutting-edge technical support for the safe and efficient use of a high proportion of renewable energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flowchart of the steps of the multi-energy coupled medium-deep buried pipe group and power grid coordinated control method of the present invention;

[0048] Figure 2 This is a logic diagram for optimizing and controlling the heating system based on model prediction in this embodiment;

[0049] Figure 3 This is a flow chart of heat distribution control between different tube groups in this embodiment;

[0050] Figure 4 This is a flow chart of the integrated energy system optimization planning scheme in this embodiment. DETAILED DESCRIPTION

[0051] The present invention will be further described below through specific examples. The illustrative examples and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0052] The method and system for coordinated control of a multi-energy coupled medium-deep buried pipe group and a power grid of the present invention include the following steps:

[0053] like Figure 1 As shown, in this embodiment, the following steps are included:

[0054] A hierarchical control architecture consisting of "bottom-level single-pipe real-time control - middle-level pipe group heat balance - top-level system planning" was designed, integrating model predictive control (MPC) with deep reinforcement learning algorithms for multi-time-scale control. The hierarchical control architecture consists of bottom, middle, and top layers. The bottom layer uses model predictive control to optimize the distribution of energy stations and satellite stations based on heat load prediction, achieving real-time and precise control of single pipes. The middle layer coordinates heat distribution between pipe groups through a heat balance algorithm. The top layer uses system planning methods for overall energy management and optimization, establishing a rolling time-domain optimization feedback mechanism.

[0055] Construct a complementary synergistic model for geothermal, wind, solar, and energy storage. Based on system dynamics and energy quality ladder matching theory, analyze the energy conversion and distribution mechanism under multiple time scales. By establishing a dynamic balance strategy for heat, electricity, and storage, effectively coordinate different energy forms, and integrate carbon trading cost prediction and electricity spot price prediction modules.

[0056] A deep reinforcement learning-based approach combined with a physical constraint embedding strategy is used to reduce the action space dimension, obtain efficiency and physical feasibility, and integrate solution tools using reinforcement learning optimization methods and efficient solvers. A multi-disturbance scenario verification platform is constructed to simulate complex operating conditions such as grid fluctuations and sudden changes in penetration rate, conduct robustness tests, and output the target model.

[0057] In this embodiment, the method for designing the bottom layer includes:

[0058] A system dispatch control method based on model prediction is adopted. The heat load demand of each heating station in the area is first predicted, and then optimization analysis and decision-making are carried out based on the prediction. The dispatch control method is categorized according to the time dimension and the space dimension. The space dimension mainly includes the equipment receiving the control command, ground source heat pumps, satellite stations, and energy stations. In the time dimension, it is divided into optimization control under short time scale and dispatch optimization under medium and long time scale according to the optimization analysis scale.

[0059] The top layer of the heating system first decomposes tasks based on energy production and orderly energy consumption plans. In the operation scheduling module, a control strategy is generated based on the specific conditions of multiple heat sources and the operating status of each device. In the intraday real-time optimization module, control targets are generated through real-time online monitoring of equipment and target optimization, and control actions are generated using advanced control methods. After completing the top-down scheduling and control process, the system collects monitoring data from lower-level equipment in real time and provides online feedback to the upper layer, realizing a closed-loop scheduling and control process. The upper layer links are corrected based on the feedback from the lower layers.

[0060] The model-based prediction-based heating system is controlled using optimized control logic. The optimized control logic includes the following steps: Future heat loads at the thermal power station level are predicted using load forecasting technology. The heating load distribution between energy stations and satellite stations is determined based on their heating capacities and heating cost curves. This maximizes the heating capacity of geothermal heat sources and coordinates efficient heating with heat sources from multiple energy stations, achieving low-carbon operation of the heating system. Based on the heat load distribution, the scheduling platform assigns heating flow rates and supply and return water temperatures to each energy station.

[0061] In the actual evaluation, the multi-unit load distribution optimization method was used to improve the overall heating efficiency and economic benefits of the units and reduce carbon emissions.

[0062] In this embodiment, the method for designing the middle layer includes:

[0063] To regulate the heat distribution between different pipe groups, the host computer platform calculates the comprehensive physical property coefficient of the terminal room, classifies heat users, determines the regulation cycle and the target return water temperature based on historical heating parameter information;

[0064] When the heat demand of the heat user changes, the return water temperature is adjusted to adapt to the heat demand. The upper computer platform issues an adjustment instruction to change the opening of the intelligent valve to change the return water temperature of the heat user to reach the target return water temperature. When the real-time return water temperature meets the target and When the heat user's intelligent valve adjustment ends, the heat user's real-time return water temperature is The target value of the real-time return water temperature of the heat user is The return water temperature deviation threshold is ξ 2h , the actual temperature of the input water is The target temperature of the input water is The temperature deviation threshold is ξ in ;

[0065] On the contrary, continue to adjust the valve opening, ξ 2h Different systems may have different thresholds. The change in the opening of the heat user's intelligent valve will cause the pressure difference and impedance of the entire network to change. At this time, the water pump of the thermal power station will perform linked frequency conversion regulation to change the flow rate and restore the pressure difference and impedance of the network to the set value. When the network impedance and pressure difference meet the requirements, the system will automatically adjust the flow rate and return to the set value. |SS op |<σ s |, and the valve opening that is most unfavorable to the heat user is the maximum limit V z =V max When the frequency conversion is ended, otherwise the frequency conversion will continue; the actual pressure difference of the pipe network is ΔP, and the set value of the pipe network pressure difference is The deviation threshold of the pipe network pressure difference is σ P The actual impedance of the pipe network is S, and the set value of the pipe network impedance is S op , the deviation threshold of the pipe network impedance is σ s .

[0066] In this embodiment, the method for designing the top layer includes:

[0067] The planning of a multi-energy coordinated integrated energy system is a conversion process between multiple energy sources. The system planning method abstracts an integrated energy system into a dual-port network of energy input and output, where multiple energy sources are converted, distributed, and stored within the integrated energy system.

[0068] Connect the input end of the system planning method to the energy network, input the corresponding electricity, gas, and oil energy at the input end, and output the energy in the form of electricity, heat, and cooling at the output end; based on the actual situation of the operation stage, establish an optimization planning model of the integrated energy system "planning and operation integration";

[0069] The optimization planning involves an optimization configuration scheme for equipment capacity and quantity, as well as optimizing the cold, heat, and power operation modes of equipment. According to the idea of decomposition and coordination, the optimization planning is transformed into a two-layer planning model. The upper-layer planning aims to maximize the internal net present value within the integrated energy system and conducts the optimization configuration of equipment capacity and quantity. The lower-layer optimization corresponds to the optimization of the operation strategy of the integrated energy system. Based on the equipment capacity and quantity optimized in the upper layer, an objective function is objectively weighted by the minimum operation cost of the integrated energy system and the maximum proportion of electric energy in the total energy consumption. According to the equipment operation constraints, the cold, heat, and power operation modes of the equipment are optimized, and the optimization objective is transmitted to the upper-layer optimization to calculate the internal rate of return. Through the iterative optimization of the upper and lower layers, an optimal "planning and operation integration" integrated energy system optimization planning scheme is obtained.

[0070] In this embodiment, the method of the integrated carbon trading cost prediction and electricity spot price prediction module includes:

[0071] On the TRNSYS platform, geothermal - wind - energy storage is coupled. The multi - energy coupled heating and cooling system model includes an energy consumption module, a cost module, and an output module. The energy consumption module is used to count the annual energy consumption of each cold and heat source subsystem. The cost module calculates the annual operation cost using time - of - use electricity prices and gas prices. The output module outputs the capacity ratio, initial investment, operation energy consumption, cooling / heating capacity, total initial investment, annual operation cost, annual carbon emissions, and life - cycle cost of each subsystem.

[0072] In the multi - energy coupled heating and cooling system model, with the help of the GenOpt optimization tool externally connected to the TrnOpt module in TRNSYS, variable optimization is carried out.

[0073] Carbon emissions are incorporated into the geothermal - wind - energy storage system, and a carbon emission cost model is established. The expression is:

[0074]

[0075] Where the basic carbon emission quota is Q0, the equivalent actual carbon emission is Q1, and the carbon emission cost is F co2 , the carbon emission cost coefficients are C1, C2, C3, C4 respectively, the first - level boundary quota of carbon emission is, and the second - level boundary quota of carbon emission is Q 02 ;

[0076] When the emission Q1 is lower than Q0, the geothermal - wind - energy storage system sells the excess emission quota. When the emission Q1 is higher than Q0, the geothermal - wind - energy storage system needs to purchase carbon emission quotas, and as the purchased quota increases, the unit purchase cost increases. When Q1 < Q0, it means that when the emission does not exceed the basic quota Q0, the excess quota is sold. When the emission exceeds different quotas, the excess part purchases carbon emission quotas according to the corresponding emission cost.

[0077] The electricity spot market conducts real-time transactions of electricity over short periods of time. The spot price reflects the dynamic balance of electricity supply and demand and is affected by multiple factors, including weather conditions, load demand, fuel prices, generation capacity, network constraints, and market participants.

[0078] Based on the advantages of different methods, a hybrid model of time series model and machine learning model, multi-model integration and decomposition-based method is proposed. The hybrid model effectively captures the complex characteristics of electricity spot prices. Among them, the hybrid of time series model and machine learning model: first use the time series model to extract the linear characteristics of price data, and then use the machine learning model to learn the nonlinear characteristics of the residual; multi-model integration: take the weighted average of the prediction results of multiple different models; decomposition-based method: first decompose the original price data into components of different frequencies, then establish a prediction model for each component, and finally combine the prediction results of each component;

[0079] By establishing a dynamic balance strategy for heat, electricity, and storage, we can achieve effective synergy among different energy forms and integrate carbon trading cost forecasting with electricity spot price forecasting modules to ensure optimal scheduling of the energy system under different market environments and operating conditions.

[0080] In actual evaluations, the hybrid model effectively captures the complex characteristics of electricity spot prices, achieving higher forecast accuracy and better generalization capabilities. By establishing a dynamic balance strategy for heat, electricity, and storage, it achieves effective synergy among different energy sources, improving overall system efficiency. By integrating carbon trading cost forecasting with electricity spot price forecasting modules, the optimization model targets economic efficiency, low carbon emissions, and stability, ensuring optimal scheduling of the energy system under different market environments and operating conditions.

[0081] In this embodiment, the method for reducing the dimension of the action space includes:

[0082] Based on the general framework of a multi-objective evolutionary algorithm for online target dimensionality reduction, the ε-MOSFLA algorithm is combined with the SORA1 and SORA2 target dimensionality reduction algorithms to construct online target dimensionality reduction algorithms that can meet different requirements. The online target dimensionality reduction algorithms include the SO1-MOSFLA algorithm and the SO2-MOSFLA algorithm. The SO1-MOSFLA algorithm uses SORA1 to reduce a fixed number of targets each time until the desired number of targets is reached, and no targets are deleted. The SO2-MOSFLA algorithm uses SORA2 to adaptively find unimportant targets and delete them according to a given error threshold until no more targets can be deleted or the number of targets is reduced to two.

[0083] Based on the online target dimensionality reduction algorithm, a target integration strategy is proposed. The targets to be deleted are classified according to their importance indicators. The importance indicators are integrated into one target through weighted accumulation and added to the target subset. This strategy solves two types of redundant and non-redundant high-dimensional multi-objective optimization problems.

[0084] This paper proposes a method based on sparse feature selection to measure the importance of the target by utilizing the geometric structure characteristics of the approximate solution set and the Pareto dominance relationship, and constructs a target dimensionality reduction algorithm for solving two different requirements.

[0085] Based on the idea of ​​sparse feature selection, a target preference ranking evaluation algorithm is proposed. When the error of the target preference ranking evaluation algorithm in the original problem does not exceed the error threshold, it starts from the original target set and only reduces one target at a time. The comprehensive ranking of the targets in each dimensionality reduction process is calculated to determine the ranking of each target.

[0086] In actual evaluation, simulation experiment results show that the target dimensionality reduction algorithm proposed in this patent can accurately delete redundant targets for high-dimensional multi-objective optimization problems with different redundancies, and its dimensionality reduction accuracy is almost unaffected by the quality of the approximate solution set, and has strong robustness.

[0087] In this embodiment, the robustness test method includes:

[0088] The labels obtained by classifying the U-shaped medium-deep buried pipe underground heat exchange model are obtained, and a local substitution model with highly similar classification capabilities to the U-shaped medium-deep buried pipe underground heat exchange model is constructed. Adversarial samples are constructed based on the output of the substitution model, and the transferability of the adversarial samples is used to achieve robustness testing of the target model.

[0089] Secondly, the multi-energy coupled medium-deep buried pipeline group and power grid coordinated control system includes:

[0090] Hierarchical multi-scale control module: This module is used to design a hierarchical control architecture consisting of "bottom-level single-tube real-time control - middle-level tube group thermal balance - top-level system planning." It integrates model predictive control (MPC) and deep reinforcement learning algorithms for multi-time-scale control. The hierarchical control architecture includes bottom, middle, and top layers. The bottom layer uses model predictive control to optimize the distribution of energy stations and satellite stations based on heat load prediction, achieving real-time and precise control of single tubes. The middle layer coordinates heat distribution between tube groups through a heat balance algorithm. The top layer uses system planning methods for overall energy management and optimization, establishing a rolling time-domain optimization feedback mechanism.

[0091] Collaborative Forecasting Module: This module is used to build a complementary synergistic model for geothermal, wind, solar, and energy storage. Based on system dynamics and energy quality ladder matching theory, it analyzes energy conversion and distribution mechanisms at multiple time scales. By establishing a dynamic balance strategy for heat, electricity, and storage, it effectively coordinates different energy forms and integrates carbon trading cost prediction and electricity spot price prediction modules.

[0092] Dimensionality reduction testing module: It is used to reduce the dimension of the action space by using a deep reinforcement learning-based combined with physical constraint embedding strategy to obtain efficiency and physical feasibility. It combines reinforcement learning optimization methods and efficient solvers to integrate solution tools, build a multi-disturbance scenario verification platform, simulate complex working conditions such as power grid fluctuations and sudden changes in penetration rate, conduct robustness tests, and output the target model.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-energy coupled coordinated control method for medium-deep buried pipe groups and power grids, characterized in that: The following steps are involved: Design a hierarchical control architecture consisting of "bottom-level single-pipe real-time control - middle-level pipe group thermal balance - top-level system planning," integrating model predictive control (MPC) with deep reinforcement learning algorithms for multi-timescale control. The hierarchical control architecture consists of a bottom layer, a middle layer, and a top layer. The bottom layer uses model predictive control to optimize the distribution of energy stations and satellite stations based on heat load forecasts, achieving real-time and precise control of single pipes. The middle layer coordinates the heat distribution between pipe groups through a heat balance algorithm. The top layer uses a system planning method to perform overall energy management and optimization, establishing a rolling time domain optimization feedback mechanism. Construct a complementary synergistic model for geothermal, wind, solar, and energy storage. Based on system dynamics and energy quality ladder matching theory, analyze the energy conversion and distribution mechanism under multiple time scales. By establishing a dynamic balance strategy for heat, electricity, and storage, effectively coordinate different energy forms, and integrate carbon trading cost prediction and electricity spot price prediction modules. A deep reinforcement learning-based approach combined with a physical constraint embedding strategy is used to reduce the dimensionality of the action space to obtain efficiency and physical feasibility. Reinforcement learning optimization methods and efficient solvers are combined to integrate solution tools, build a multi-disturbance scenario verification platform, simulate complex working conditions such as grid fluctuations and sudden changes in penetration rate, conduct robustness tests, and output control results.

2. The method for coordinated control of a multi-energy coupled medium-deep buried pipe group and a power grid according to claim 1 is characterized in that: The method for designing the bottom layer includes: A system dispatch control method based on model prediction is adopted. The heat load demand of each heating station in the area is first predicted, and then optimization analysis and decision-making are carried out based on the prediction. The dispatch control method is categorized according to the time dimension and the space dimension. The space dimension mainly includes the equipment receiving the control command, ground source heat pumps, satellite stations, and energy stations. In the time dimension, it is divided into optimization control under short time scale and dispatch optimization under medium and long time scale according to the optimization analysis scale. The top layer of the heating system first decomposes tasks based on energy production and orderly energy consumption plans. In the operation scheduling module, a control strategy is generated based on the specific conditions of multiple heat sources and the operating status of each device. In the intraday real-time optimization module, control targets are generated through real-time online monitoring of equipment and target optimization, and control actions are generated using advanced control methods. After completing the top-down scheduling and control process, the system collects monitoring data from lower-level equipment in real time and provides online feedback to the upper layer, realizing a closed-loop scheduling and control process. The upper layer links are corrected based on the feedback from the lower layers. The heating system based on model prediction is controlled by optimized control logic, in which the optimized control logic is as follows: the future heat load at the thermal power station level is predicted through load forecasting technology, and the heating load distribution of the energy station and satellite station is determined in combination with the heating capacity and heating cost curve of the energy station and satellite station. Under the premise of maximizing the heating capacity of geothermal heat sources, efficient heating is coordinated with the heat sources of multiple energy stations to achieve low-carbon operation of the heating system; based on the heat load distribution, the scheduling platform gives the heating flow and supply and return water temperature of each energy station.

3. The method for coordinated control of a multi-energy coupled medium-deep buried pipeline group and a power grid according to claim 1 is characterized in that: The method for designing the middle layer includes: To regulate the heat distribution between different pipe groups, the host computer platform calculates the comprehensive physical property coefficient of the room, classifies heat users, determines the regulation period and the target return water temperature for the terminal based on historical heating parameter information. When the heat demand of the heat user changes, the return water temperature is adjusted to adapt to the heat demand. The upper computer platform issues an adjustment instruction to change the opening of the intelligent valve to change the return water temperature of the heat user to reach the target return water temperature. When the real-time return water temperature meets the target and When the heat user's intelligent valve adjustment ends, the heat user's real-time return water temperature is The target value of the real-time return water temperature of the heat user is The return water temperature deviation threshold is ξ 2h , the actual temperature of the input water is The target temperature of the input water is The temperature deviation threshold is ξ in ; On the contrary, continue to adjust the valve opening, ξ 2h Different systems may have different thresholds. The change in the opening of the heat user's intelligent valve will cause the pressure difference and impedance of the entire network to change. At this time, the water pump of the thermal power station will perform linked frequency conversion regulation to change the flow rate and restore the pressure difference and impedance of the network to the set value. When the network impedance and pressure difference meet the requirements, the system will automatically adjust the flow rate and return to the set value. |SS op |<σ s |, and the valve opening that is most unfavorable to the heat user is the maximum limit V z =V max When the frequency conversion is ended, otherwise the frequency conversion will continue; the actual pressure difference of the pipe network is ΔP, and the set value of the pipe network pressure difference is The deviation threshold of the pipe network pressure difference is σ P The actual impedance of the pipe network is S, and the set value of the pipe network impedance is S op , the deviation threshold of the pipe network impedance is σ s .

4. The method for coordinated control of a multi-energy coupled medium-deep buried pipeline group and a power grid according to claim 1 is characterized in that: The method for designing the top layer includes: The planning of a multi-energy collaborative integrated energy system is a conversion process between multiple energies. The system planning method abstracts an integrated energy system into an input-output dual-port network of energy. Multiple energies are converted, distributed, and stored within the integrated energy system. Connect the input end of the system planning method to the energy network, input corresponding electrical, gas, and oil energies at the input end, and output energies in the forms of electricity, heat, and cold at the output end. According to the actual situation in the operation stage, establish an "integration of planning and operation" optimization planning model for the integrated energy system. The optimization planning involves an optimized configuration plan for the equipment capacity and quantity, as well as optimizing the cold, heat, and electricity operation modes of the equipment. According to the idea of decomposition and coordination, the optimization planning is transformed into a two-layer planning model. The upper-layer planning aims to maximize the net present value within the integrated energy system and conducts the optimized configuration of the equipment capacity and quantity. The lower-layer optimization corresponds to the optimization of the operation strategy of the integrated energy system. Based on the equipment capacity and quantity optimized in the upper layer, an objective function is objectively weighted by the minimum value of the operation cost of the integrated energy system and the maximum value of the proportion of electricity in the total energy consumption. According to the equipment operation constraints, the cold, heat, and electricity operation modes of the equipment are optimized, and the optimization target is transmitted to the upper-layer optimization to calculate the internal rate of return. Through the iterative optimization of the upper and lower layers, an optimal "integration of planning and operation" integrated energy system optimization planning scheme is obtained.

5. The method for coordinated control of a multi-energy coupled medium-deep buried pipeline group and a power grid according to claim 1 is characterized in that: The method for integrating the carbon trading cost prediction and electricity spot price prediction module includes: Couple geothermal, wind-solar, and energy storage on the TRNSYS platform. The multi-energy coupled heating and cooling system model includes an energy consumption module, a cost module, and an output module. Among them, the energy consumption module is used to count the annual energy consumption of each cold and heat source subsystem; the cost module calculates the annual operation cost using time-of-use electricity prices and gas prices; the output module outputs the capacity ratio, initial investment, operation energy consumption, cooling / heating capacity, total initial investment, annual operation cost, annual carbon emissions, and life cycle cost of each subsystem. In the multi-energy coupled heating and cooling system model, use the GenOpt optimization tool externally connected to the TrnOpt module in TRNSYS to optimize variables. Incorporate carbon emissions into the geothermal, wind-solar, and energy storage system and establish a carbon emission cost model, with the expression: The carbon emission base is Q0, the equivalent actual carbon emission is Q1, and the carbon emission cost is F co2 , the carbon emission cost coefficients are C1, C2, C3, and C4 respectively, and the first-level boundary quota of carbon emissions is Q 01 , the secondary boundary quota of carbon emissions is Q 02 ; When the emission amount Q1 is lower than Q0, the geothermal, wind-solar, and energy storage system sells the excess emission allowances; when the emission amount Q1 is higher than Q0, the geothermal, wind-solar, and energy storage system needs to purchase carbon emission allowances, and as the purchased amount increases, the unit purchase cost increases; when Q1 < Q0, it means that when the emissions do not exceed the base amount Q0, the excess amount is sold; when the emissions exceed different amounts, the excess part purchases carbon emission allowances according to the corresponding emission costs. The electricity spot market conducts real-time transactions of electricity over short periods of time. The spot price reflects the dynamic balance of electricity supply and demand and is affected by multiple factors, including weather conditions, load demand, fuel prices, generation capacity, network constraints, and market participants. Based on the advantages of different methods, a hybrid model of time series model and machine learning model, multi-model integration and decomposition-based method is proposed. The hybrid model effectively captures the complex characteristics of electricity spot prices. Among them, the hybrid of time series model and machine learning model: first use the time series model to extract the linear characteristics of price data, and then use the machine learning model to learn the nonlinear characteristics of the residual; multi-model integration: take the weighted average of the prediction results of multiple different models; decomposition-based method: first decompose the original price data into components of different frequencies, then establish a prediction model for each component, and finally combine the prediction results of each component; By establishing a dynamic balance strategy for heat, electricity and storage, effective coordination between different energy forms can be achieved, and the carbon trading cost forecasting and electricity spot price forecasting modules are integrated to ensure optimal scheduling of the energy system under different market environments and operating conditions.

6. The method for coordinated control of a multi-energy coupled medium-deep buried pipeline group and a power grid according to claim 1, characterized in that: The method for reducing the dimension of the action space includes: Based on the general framework of a multi-objective evolutionary algorithm for online target dimensionality reduction, the ε-MOSFLA algorithm is combined with the SORA1 and SORA2 target dimensionality reduction algorithms to construct online target dimensionality reduction algorithms that can meet different requirements. The online target dimensionality reduction algorithms include the SO1-MOSFLA algorithm and the SO2-MOSFLA algorithm. The SO1-MOSFLA algorithm uses SORA1 to reduce a fixed number of targets each time until the desired number of targets is reached, and no targets are deleted. The SO2-MOSFLA algorithm uses SORA2 to adaptively find unimportant targets and delete them according to a given error threshold until no more targets can be deleted or the number of targets is reduced to two. Based on the online target dimensionality reduction algorithm, a target integration strategy is proposed. The targets to be deleted are classified according to their importance indicators. The importance indicators are integrated into one target through weighted accumulation and added to the target subset. This strategy solves two types of redundant and non-redundant high-dimensional multi-objective optimization problems. This paper proposes a method based on sparse feature selection to measure the importance of the target by utilizing the geometric structure characteristics of the approximate solution set and the Pareto dominance relationship, and constructs a target dimensionality reduction algorithm for solving two different requirements. Based on the idea of ​​sparse feature selection, a target preference ranking evaluation algorithm is proposed. When the error of the original problem does not exceed the error threshold, the target preference ranking evaluation algorithm starts from the original target set and only reduces one target at a time. The comprehensive ranking of the targets in terms of importance in each dimensionality reduction process is statistically analyzed to determine the ranking of each target.

7. The method for coordinated control of a multi-energy coupled medium-deep buried pipeline group and a power grid according to claim 1, characterized in that: The robustness testing method comprises: The labels obtained by classifying the U-shaped medium-deep buried pipe underground heat exchange model are obtained, and a local substitution model with highly similar classification capabilities to the U-shaped medium-deep buried pipe underground heat exchange model is constructed. Adversarial samples are constructed based on the output of the substitution model, and the transferability of the adversarial samples is used to achieve robustness testing of the target model.

8. A multi-energy coupled medium-deep buried pipe group and power grid coordinated control system for executing the method according to any one of claims 1 to 7, characterized in that: include: Hierarchical multi-scale control module: This module is used to design a hierarchical control architecture consisting of "bottom-level single-pipe real-time control - middle-level pipe group thermal balance - top-level system planning." It integrates model predictive control (MPC) with deep reinforcement learning algorithms for multi-timescale control. The hierarchical control architecture consists of a bottom layer, a middle layer, and a top layer. The bottom layer uses model predictive control to optimize the distribution of energy stations and satellite stations based on heat load prediction, achieving real-time and precise control of single pipes. The middle layer coordinates the heat distribution between pipe groups through a heat balance algorithm. The top layer uses a system planning method to conduct overall energy management and optimization, establishing a rolling time domain optimization feedback mechanism. Collaborative Forecasting Module: This module is used to build a complementary synergistic model for geothermal, wind, solar, and energy storage. Based on system dynamics and energy quality ladder matching theory, it analyzes energy conversion and distribution mechanisms at multiple time scales. By establishing a dynamic balance strategy for heat, electricity, and storage, it effectively coordinates different energy forms and integrates carbon trading cost prediction and electricity spot price prediction modules. Dimensionality reduction testing module: It is used to reduce the dimension of the action space by using a deep reinforcement learning-based combined with physical constraint embedding strategy to obtain efficiency and physical feasibility. It combines reinforcement learning optimization methods and efficient solvers to integrate solution tools, build a multi-disturbance scenario verification platform, simulate complex working conditions such as power grid fluctuations and sudden changes in penetration rate, conduct robustness tests, and output the target model.

Images