IES multi-time scale composite control method and system based on dual feedback RMPC

Abstract

The present invention belongs to the technical field of integrated energy systems, and provides an IES multi-time scale composite control method and system based on dual-feedback RMPC, analyzes the multi-time scale characteristics of typical equipment and IDR, establishes a composite control framework for equipment side-load side coordination, and performs differentiated control on different time scales through a three-layer controller within a day in view of the differences in regulation rates and response characteristics between multi-energy equipment and IDR, modifies the upper-layer reference layer by layer, and issues control strategies in real time, fully stimulating the flexibility of both the supply and demand sides.

Description

Technical Field

[0001] The present invention belongs to the technical field of integrated energy systems, and in particular to an IES multi-time scale composite control method and system based on dual-feedback robust model predictive control (RMPC). Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] With the continuous increase in the penetration rate of renewable energy and the rapid popularization of energy market-oriented reforms, the uncertainty on both the supply and demand sides of the integrated energy system (IES) is rapidly increasing, and its efficient and reliable operation is facing challenges.

[0004] In order to overcome the external uncertain environment from the root, some researchers have tried to improve the prediction accuracy by modifying the prediction model. However, the inherent randomness and volatility of the source load means that the prediction error can never be completely eliminated. Although the improved prediction model has alleviated the impact of uncertain factors to a certain extent, it still does not touch the core problem and the uncertain environment still exists.

[0005] Advanced scheduling strategies are another key to meeting the challenges of uncertain environments. Especially when accurate prediction results cannot be obtained, scheduling strategies that take into account the impact of uncertainty become increasingly important. The advanced scheduling strategies currently proposed include a series of advanced scheduling strategies including stochastic optimization, robust optimization, fuzzy chance constrained programming, multi-time scale optimization, model predictive control (MPC) and its derivative methods. These studies have provided valuable experience for energy management in different scenarios and alleviated the adverse effects of uncertain factors. However, the above methods all have certain defects when applied alone, such as poor economy of robust optimization, poor reliability of stochastic optimization, insufficient accuracy of prediction models, poor anti-interference ability of multi-time scale optimization, and MPC's inability to adapt to multiple uncertain environments. Existing scheduling strategies still cannot help IES overcome uncertain environments.

[0006] Demand response (DR) is the key to stimulate the interaction between demand-side resources and renewable energy. Secondary dispatch of loads through DR is of great significance to mobilize the demand side to resist uncertainty and improve the absorption capacity of renewable energy. In particular, DR in IES has developed into integrated demand response (IDR) covering more load types. Compared with DR, it has higher flexibility and is more critical for guiding horizontal multi-energy complementarity and vertical source-load coordination in the system. However, due to the strong interaction and dynamic characteristics differences between multiple heterogeneous energy sources, it is difficult for existing methods to reasonably give the response of different IDRs in different periods, and the flexibility is insufficient, which cannot give full play to the complementary characteristics of IDR and renewable energy.

[0007] In summary, domestic and foreign scholars have proposed a variety of solutions for the efficient operation of IES in uncertain environments, trying to guide IES to give full play to its advantages of multi-energy complementarity and source-load coordination through accurate prediction models or advanced scheduling strategies and new IDRs to weaken the impact of uncertain environments. However, at present, the above solutions all have certain difficulties in implementation, and there is still no perfect way to solve the impact of uncertainty. At present, there is an urgent need for an effective integrated scheduling framework to coordinate the existing methods, maximize their advantages, and eliminate their shortcomings. Summary of the invention

[0008] In order to solve at least one technical problem existing in the above-mentioned background technology, the present invention provides an IES multi-time scale composite control method and system based on dual-feedback robust model predictive control (RMPC), which organically integrates the multi-time scale optimization idea, robust optimization algorithm, advanced prediction model, IDR strategy and MPC itself by leveraging the flexibility and scalability advantages of the MPC framework, and establishes a composite control framework based on dual-feedback RMPC, which innovatively realizes the coordination of the source-load prediction model and the robust optimization algorithm through the outer loop real-time feedback, successfully achieves dynamic adaptive adjustment of uncertainty, and draws on the source-load uncertainty processing method, utilizes the inner loop real-time feedback to model the uncertainty of equipment output disturbance, and enhances the reliability of the control sequence.

[0009] In order to achieve the above object, the present invention adopts the following technical solution:

[0010] The first aspect of the present invention provides an IES multi-time scale composite control method based on dual feedback RMPC, comprising the following steps:

[0011] Obtain day-ahead source-load forecast data for integrated energy systems;

[0012] Based on the acquired day-ahead source-load forecast data of the integrated energy system, the electric load incentive demand response and the output of related power equipment are obtained through two stages: day-ahead dispatch and intraday correction.

[0013] Among them, in the day-ahead dispatch stage, the initial reference dispatch plan for the price-based demand response of electrical loads and electricity, heat and gas related equipment is determined;

[0014] In the intra-day correction stage, based on the price-based demand response of electrical loads and the initial reference scheduling plan for electricity, heat and gas related equipment, the intra-day correction is divided into fast, intermediate and slow control sub-layers according to the differences in the dynamic characteristics of multi-heterogeneous energy sources. Controllers with different time scales are used for differentiated control. The control time interval of each layer is used as the corresponding scale. A double-layer feedback architecture is constructed at each layer. The output disturbance of the equipment and demand response obtained in the inner loop is fed back to the outer loop. In the outer loop, the source and load data are updated, the system state prediction model is dynamically trained, and the robust optimization algorithm is guided to adjust the uncertain parameters of the system. The reference scheduling plan output by the previous layer is adjusted layer by layer, and the output of electrical load incentive demand response and related power equipment is output in the last layer.

[0015] A second aspect of the present invention provides an IES composite control system based on dual feedback RMPC, comprising:

[0016] The day-ahead source-load forecast data acquisition module is configured to: acquire day-ahead source-load forecast data of the integrated energy system;

[0017] The initial dispatch plan acquisition module is configured to: obtain the initial reference dispatch plan of the electric load price-type demand response and the electric, heat and gas related equipment through day-ahead dispatch based on the acquired day-ahead source-load forecast data of the integrated energy system;

[0018] The composite control module is configured as follows:

[0019] Based on the acquired day-ahead source-load forecast data of the integrated energy system, the electric load incentive demand response and the output of related power equipment are obtained through two stages: day-ahead dispatch and intraday correction.

[0020] Among them, in the day-ahead dispatch stage, the initial reference dispatch plan for the price-based demand response of electrical loads and electricity, heat and gas related equipment is determined;

[0021] In the intra-day correction stage, based on the price-based demand response of electrical loads and the initial reference scheduling plan for electricity, heat and gas related equipment, the intra-day correction is divided into fast, intermediate and slow control sub-layers according to the differences in the dynamic characteristics of multi-heterogeneous energy sources. Controllers with different time scales are used for differentiated control. The control time interval of each layer is used as the corresponding scale. A double-layer feedback architecture is constructed at each layer. The output disturbance of the equipment and demand response obtained in the inner loop is fed back to the outer loop. In the outer loop, the source and load data are updated, the system state prediction model is dynamically trained, and the robust optimization algorithm is guided to adjust the uncertain parameters of the system. The reference scheduling plan output by the previous layer is adjusted layer by layer, and the output of electrical load incentive demand response and related power equipment is output in the last layer.

[0022] A third aspect of the present invention provides a computer-readable storage medium.

[0023] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the IES multi-time-scale composite control method based on dual-feedback RMPC as described above.

[0024] A fourth aspect of the present invention provides a computer device.

[0025] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the IES multi-time scale composite control method based on dual feedback RMPC as described above are implemented.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The present invention organically integrates the multi-time scale optimization concept, robust optimization algorithm, advanced prediction model, IDR strategy and MPC itself by taking advantage of the flexibility and scalability of the MPC framework, and innovatively establishes a composite control framework based on dual-feedback RMPC. The coordination of the source-load prediction model and the robust optimization algorithm is achieved through the outer loop real-time feedback, and the dynamic adaptive adjustment of uncertainty is successfully achieved. In addition, the uncertainty processing method of the source-load is referenced, and the uncertainty of the output disturbance of the modeling equipment is modeled by the inner loop real-time feedback, thereby enhancing the reliability of the control sequence. Through the coordination of various components, the present invention can fully release the adjustment potential of the flexible resources on both sides of the source and load, and resist the interference of the external uncertain environment.

[0028] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0030] Figure 1 is a structural diagram of an IES provided by an embodiment of the present invention;

[0031] Figure 2 is a dual-loop feedback RMPC architecture diagram provided by an embodiment of the present invention;

[0032] Figure 3 It is a source-load composite control framework diagram of a dual-loop feedback RMPC provided by an embodiment of the present invention;

[0033] Figure 4 The multi-time scale prediction result of the electric load provided by the embodiment of the present invention;

[0034] Figure 5 The multi-time scale prediction result of gas load provided by the embodiment of the present invention;

[0035] Figure 6 The multi-time scale prediction result of heat load provided by the embodiment of the present invention;

[0036] Figure 7 The day-ahead dispatch result of electric power provided by the embodiment of the present invention;

[0037] Figure 8 The gas power day-ahead scheduling result provided by the embodiment of the present invention;

[0038] Fig. 9 The thermal power day-ahead scheduling result provided by the embodiment of the present invention;

[0039] Fig.10 The thermal power 1-hour scheduling result provided by the embodiment of the present invention;

[0040] Fig.11 The 15-minute scheduling result of gas power within a day provided by the embodiment of the present invention;

[0041] Fig.12 The 5-minute scheduling result of electric power within a day provided by the embodiment of the present invention;

[0042] Fig.13 The 15-minute adjustment result of the natural gas-related equipment provided in the embodiment of the present invention;

[0043] Fig.14 The day-ahead comprehensive demand response result provided by the embodiment of the present invention;

[0044] Fig.15 The 1-hour comprehensive demand response result provided by the embodiment of the present invention;

[0045] Fig.16 The comprehensive demand response result within 15 minutes within a day provided by the embodiment of the present invention;

[0046] Fig.17 This is the 5-minute comprehensive demand response result within a day provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0047] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0048] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0049] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0050] Embodiment 1

[0051] like Figure 1 As shown, the IES includes three energy subsystems, namely, an electric power subsystem, a natural gas subsystem and a thermal subsystem. The subsystems are interconnected and coupled. The electric power subsystem includes a wind turbine, a photovoltaic device and a power storage device; the natural gas subsystem includes a power-to-gas device and a gas storage device; the thermal subsystem includes a cogeneration device, a gas boiler, an electric boiler and a heat storage device.

[0052] The upstream nodes of IES are connected to the external power grid and natural gas grid, and are internally coupled with energy supply, conversion and storage equipment in various energy forms; the downstream nodes cover the three energy demands of electricity, gas and heat and IDR.

[0053] Among them, IDR includes price-based, substitution-based and incentive-based DR; energy conversion equipment includes power-to-gas system (P2G), combined heat and power system (CHP), electric boiler (EB) and gas boiler (GB). Energy storage equipment includes hot water storage tank (HS), gas storage tank (GS) and battery (ES). The system is equipped with wind turbines (WT) and photovoltaics (PV) to input renewable energy into the energy network. End users can obtain diversified energy supply services from IES according to their actual energy needs.

[0054] On the device side:

[0055] The electricity, gas, and heat subsystems have significant multi-time scale characteristics. A unified dispatch cycle cannot take into account the control performance of each subsystem. In the power subsystem, the power grid can respond instantly to dispatch instructions and immediately reach a steady-state value, allowing related equipment to respond to instantaneous power fluctuations by receiving higher-frequency control instructions without being constrained by energy transmission. As for the heat and gas subsystems, due to the certain thermal inertia and pipe storage characteristics of the heat network and building complexes, the response process after the dispatch instruction is issued requires a certain amount of dynamic change to reach the steady-state value. Frequent adjustment of heat and gas equipment will cause the system state to be in a transient change process at all times, destroying the energy supply quality.

[0056] The determination of the dispatch cycle should take into account the transmission characteristics of multiple heterogeneous energy flows and the control characteristics of multiple types of equipment. Even if some equipment has a short response time, it is not suitable to adjust frequently due to energy transmission limitations. For example, in the combined heat and power system CHP, the core equipment gas turbine has a fast response rate, but the output involves thermal energy and is connected to the heat network. Due to the influence of thermal inertia and heat medium transmission limitations, the dispatch cycle should not be too short. Typical equipment adjustment rates in IES are as follows: Figure 4 As shown. Figure 4 It can be seen that the operating characteristics of equipment with power electronic devices as core controllers and combustion equipment are significantly different. When these units are operated together, it is likely that the system control deviation will occur due to the differences in their own response characteristics and the dynamic characteristics of the energy network. To this end, a multi-time scale control strategy for the production capacity side coordinated by each subsystem is established.

[0057] On the demand side

[0058] The multi-time scale characteristics of IDR are concentrated in two aspects. First, due to different pricing methods and incentive mechanisms, PBDR and IBDR belong to response resources in different periods. Among them, the execution of PBDR is that users spontaneously adjust their energy consumption behavior according to the price information released by energy companies, and the response time is long. In addition, energy prices generally adopt the day-ahead pricing model, so PBDR needs to respond on the day before. The execution of IBDR is that energy companies sign contracts with users in advance and stipulate the response rate according to the contract. When necessary, load control can be directly carried out, and the response time is short. Therefore, IBDR is not limited to day-ahead response. It can be divided into three types according to the time of advance notice: Class A IBDR (day-ahead response), Class B IBDR (15min-2h response within the day), Class C IBDR (5min-15min within the day) and Class D IBDR (real-time response). Second, IDR covers response resources of more load types and is affected by the response characteristics of multiple heterogeneous energy flows. Among them, ADR involves the mutual conversion of electricity, gas and heat. Limited by the transmission characteristics of different energy sources and the response characteristics of coupled equipment, there are technical barriers to unified execution at the same time scale. In addition, in IES, multiple types of loads can execute IBDR. However, due to energy characteristics, different loads have different execution periods. For example, gas loads cannot participate in real-time response within the day due to the pipe storage characteristics of the gas grid. Obviously, if all DR resources are uniformly scheduled, it will be difficult to fully tap the regulation potential of load-side resources and absorb new energy with intraday deviations. For this reason, an IDR multi-time scale coordinated control strategy is established.

[0059] Based on the multi-time scale response characteristics of the equipment side and the demand side and the RMPC dual closed-loop feedback mechanism, the present invention proposes a source-load multi-time scale composite control strategy, which is divided into two stages: day-ahead scheduling and intraday correction. The day-ahead optimization takes 24 hours as the scheduling cycle, formulates the unit start-stop plan and equipment operation plan for 24 time periods as the reference of the intraday correction layer to ensure the global performance of the control strategy, and determines the PBDR and Class A IBDR call plan. In view of the lack of anti-interference ability and robustness of open-loop scheduling, the RMPC method is introduced in the intraday correction layer. The RMPC method is based on the idea of ​​rolling time domain, combines the current operating state of the system with the predicted state in the future, obtains the optimal control sequence, and uses the first item of the control sequence for the actual control of the system. Then, the above control process is repeated. According to the differences in the dynamic characteristics of multiple heterogeneous energy sources, the intraday correction layer is divided into fast, intermediate and slow control sub-layers, and RMPC controllers of different time scales are used for differentiated control to absorb source-load uncertainty. Among them, the slow control sublayer prediction time domain is 4h and the time scale is 1h. It is used to control the thermal subsystem with slow dynamic characteristics, adjust the output of CHP, GB, EB, and HS units, and determine the electric heat ADR call plan.

[0060] In order to further reduce the impact of uncertainty in the output of renewable energy, multi-energy loads and controlled objects, the slow control sublayer, as a bridge connecting different control stages, adjusts the reference plans of the electricity and gas subsystems on the day before while controlling the thermal subsystem, and issues a reference that is more in line with the actual situation and has globality. The intermediate control sublayer has a prediction time domain of 2 hours and a time scale of 15 minutes. It is used to control the natural gas subsystem, adjust the gas purchase volume, P2G, and GS unit output, and determine the electrical ADR and Class B IBDR call plan, and adjust the power subsystem reference plan. The fast control sublayer has a prediction time domain of 20 minutes and a time scale of 5 minutes. It is used to control the power subsystem, adjust the purchased gas volume, ES unit output, and determine the Class C IBDR call plan.

[0061] Based on the conventional MPC strategy, the present invention establishes a double closed-loop feedback mechanism such as Figure 2 The inner layer is the actual output and output disturbance feedback of the equipment and demand response, and the outer layer is the real-time source and load data and prediction model training feedback.

[0062] The inner layer can effectively avoid the accumulation of deviations between the dispatch value and the actual output during the rolling solution process by real-time monitoring and collecting the operating status of the equipment and DR. At the same time, the uncertainty of equipment and DR output disturbance is modeled, and the incremental state space equation containing equipment and DR output disturbance terms is established under the RMPC framework to enhance the reliability of the control strategy.

[0063] Considering the uncertainty of the actual controlled object of the system, the planned output and the actual output of the equipment and DR cannot be exactly the same. After the control quantity implements the control effect, there will be a deviation between the actual output value of the controlled object and the planned value. Therefore, this embodiment draws on the processing method of renewable energy and multi-energy load uncertainty, and sets δ as the equipment output disturbance, which is used as the robust optimization uncertainty and described as follows using the cardinality uncertainty set:

[0064]

[0065] in, δ are the upper limits of the equipment disturbance error, is the actual value of the disturbance error of the device at the previous moment, obtained through feedback. The disturbance that may be generated by the device and DR is simulated using δ, and is incorporated into the state space model at time t+1 using the inner loop feedback, and is used as a constraint to obtain a more reliable control strategy.

[0066] Taking the real-time control layer as an example, a system state prediction model based on the state space equation is established:

[0067]

[0068] In the formula, the state variable matrix of equipment output x(t) = [P ES.chr( t), P ES.dis (t), S ES (t), P grid (t), IBDR III (t)] T , considering the control variable matrix of output disturbance u(k)⊙d(k)=ΔP ES.che (t)(1+ξ ES.chr (t)), ΔP ES.dis (t)(1+ξ ES.dis (t)), ΔIBDR(t) III (1+ξ III (t)) T , d(t) is the disturbance of the device and IDR output, the disturbance variable matrix Output variable matrix y c (t) = [S ES (t), P grid (t)] T ;

[0069] set up According to the IES real-time power balance and energy storage SOC iteration equation, we can get:

[0070]

[0071] Among them, ΔP ES.che (t), ΔP ES.dis (t) are the output changes of the power storage device relative to the previous moment, ΔIBDR(t) III is the output change of the three types of incentive demand response of the electric load relative to the previous moment, ΔP e，load (t), ΔP wind (t), ΔP pv (t) are the output of electric load, wind power and photovoltaic power relative to the previous moment, are the upper layer reference plan values ​​of MT, P2G and EB devices at time t, ξ ES.chr (t), ξ ES.dis (t), ξ III (t) are the battery charge and discharge and IBDR output disturbance at time t, S ES (t), P grid (t) are the battery storage energy and the power purchased from the superior power grid at time t.

[0072] The outer layer dynamically trains the prediction model by updating the real data of the source and load in real time, and guides the uncertainty adjustment of the robust optimization algorithm. In this regard, the present invention innovatively designs a synergistic mechanism between the prediction model training evaluation index and the robust optimization uncertainty parameter, which simply and effectively achieves a dynamic balance between the conservatism and economy of the algorithm.

[0073] The synergistic approach of predictive model accuracy and robust optimization of uncertain parameters is as follows:

[0074] The description method of the uncertain parameters in robust optimization is to select the cardinality type uncertainty set, which can be expressed as:

[0075]

[0076] Where ξ is the actual value of the uncertain parameters, including wind power, photovoltaic output, and electricity, heat, and gas loads. is the predicted value of the uncertain parameter, ξ They are the upper and lower bounds of the prediction error, which can be set manually. Γ is the uncertainty parameter, and its value range is 0-1.

[0077] The mean absolute percentage error (MAPE) is selected as the evaluation index of the prediction model, which can be expressed as:

[0078]

[0079] Where n is the time step of the solution domain.

[0080] According to the cardinality uncertainty set and MAPE expression, ξ can be expressed in the following two forms:

[0081]

[0082]

[0083] Just order The two forms of ξ value representation are basically the same, and for the MAPE expression, when the prediction model MAPE is known, the value range of ξ can be basically determined.

[0084] Obviously, it is feasible to guide robust optimization by MAPE, which can avoid the defects of the algorithm being too conservative or unreliable due to the random setting of Γ. It should be noted that the value range of Γ is 0-1, while according to the existing advanced prediction models, the value range of MAPE is [0, MAPE max ], MAPE maxIt is generally 5% (the prediction accuracy of wind power, photovoltaic output, and electricity, gas, and heat loads are different, and the electricity load is used as an example here), and MAPE is the average prediction error. In order to ensure the reliability of the solution, considering some extreme cases, the MAPE value range is expanded to [0, 2*MAPE max ].

[0085] Since the two have different value ranges, this embodiment makes You can directly Control the conservatism of robust optimization. If the prediction error of the training model is larger, the robust optimization algorithm Γ will be larger, and the worst source-load output scenario will be more severe. At this time, the control strategy obtained will be more robust, solving the problem of inaccurate prediction.

[0086] The source-load composite control framework based on the dual-loop feedback RMPC proposed in the present invention is as follows: Figure 3 By analyzing the significant multi-time scale characteristics among multiple heterogeneous energy flows in IES, a multi-time scale control framework for equipment-load coordination is developed, and the scheduling periods of different equipment and DR are given.

[0087] After the day-ahead dispatch, the global initial reference dispatch plan is obtained. In the intraday stage, the planned values ​​are adjusted layer by layer to make it closer to the actual situation, and the output plans of heating, gas and electricity equipment are issued in turn. Specifically, in the day-ahead dispatch stage, the price-based demand response (PBDR) of the electric load and the initial reference plan of the electric heat-related equipment are obtained through the day-ahead source and load forecast data; in the intraday one-hour dispatch stage, the electric heat alternative demand response (ADR), the incentive based demand response (IBDR) of the electric and gas loads and the output of the CHP, GB, EB, HS equipment are obtained according to the latest source and load forecast data within one hour of the day, and the reference plan of electricity and gas for the fifteen-minute dispatch period is given.

[0088] In the 15-minute dispatching stage, the ADR of electricity and gas, the B-type IBDR of electricity and gas loads, the output of P2G and GS equipment, and the amount of gas purchased are obtained based on the latest source-load forecast data within 15 minutes of the day, and the electricity reference plan for the five-minute dispatching period is given. In the five-minute dispatching stage, the C-type IBDR of electricity loads and the output of ES, WT, PV equipment and the amount of electricity purchased are obtained.

[0089] The objective function consists of two parts: day-ahead and intra-day. Day-ahead mainly considers the economic efficiency of system operation, so the day-ahead objective function is expressed as:

[0090]

[0091] Where F is the total operating cost of IES, N is the scheduling period. In this embodiment, the time interval is 1 hour. is the energy purchase cost of the i-th type of energy, is the operation and maintenance cost of the i-th type of equipment, is the demand response cost of the i-th type.

[0092] The IES electricity purchase cost and gas purchase cost are proportional to its purchased electricity power and gas purchase power respectively, so they can be expressed as:

[0093]

[0094]

[0095] In the formula, is the power purchased by IES from the distribution grid, is the gas purchasing power of IES, is the electricity price purchased from the distribution network at time t, is the natural gas price at time t.

[0096] The demand response cost of IES can be summarized as:

[0097]

[0098] In the formula, is the load variation of the i-th type, is the system's compensation cost for the i-th type of load change.

[0099] The objective function of the intraday rolling part includes three parts: the objective function of the power subsystem controller, with a control interval of 1 hour; the objective function of the natural gas subsystem controller, with a control interval of 15 minutes; and the objective function of the thermal subsystem controller, with a control interval of 5 minutes.

[0100] The objective functions of the electrical, gas, and thermal subsystems can be uniformly described as:

[0101]

[0102] In the formula, F i is the scheduling cost of the ith subsystem, is the output of the i-th type of equipment at time t within a day, is the output of the i-th type of equipment at time t-1 within the day, μ i is the output adjustment cost of the i-th type of equipment, is the reference scheduling plan for the i-th type of equipment at time t, is the output change penalty for the i-th type of equipment.

[0103] The constraints include:

[0104] (1) Equipment output constraints, including energy conversion equipment groups, renewable energy units, hybrid energy storage systems, and power upper and lower limits, specifically:

[0105] Energy conversion equipment group

[0106] The energy conversion equipment in the system includes CHP, P2G, EB, and GB. The input and output power balance equation is as follows:

[0107]

[0108] In the formula, are the electrical power and thermal power output of the CHP unit respectively; η CHP,e , η CHP,h They are the electricity and gas production efficiencies of the CHP unit; Input gas power for CHP unit; are the output gas power and input electric power of P2G, respectively, P2G is the power-to-gas efficiency; are the output thermal power and input electrical power of EB, respectively, EB is the heat generation efficiency of EB; are the output heat power and input gas power of GB respectively, η GB is the heat generation efficiency of GB.

[0109] Renewable energy units

[0110] The system is equipped with WT and PV to input renewable energy into the energy grid. The output power of WT and PV are as follows:

[0111]

[0112] in, is the power generated by WT at time t, v in and v out are the cut-in wind speed and cut-out wind speed of WT, v e is the rated wind speed, is the rated power of the fan.

[0113]

[0114] in, is the electric power output by PV at time t, is the rated power of PV, G t and G ref are the light intensity at time t and the reference light intensity under standard test conditions, T0 and T ref are ambient temperature and reference temperature under standard test conditions, respectively.

[0115]

[0116] in, are the cooling power and thermal power consumed by the absorption refrigerator at time t, η AC is the refrigeration coefficient of the absorption chiller.

[0117] Hybrid energy storage system:

[0118]

[0119] In the formula, i represents the load type, which here represents electricity, heat, and gas loads. is the energy storage at time t, δ i is the self-loss coefficient, and are the charging and discharging power at time t, η i,ch and η i,dis are the charging and discharging energy efficiencies respectively.

[0120] Power upper and lower limits:

[0121]

[0122] In the formula, i represents the energy conversion device, electricity purchase, gas purchase and charging and releasing energy status, is the output power of the i-th type of equipment at time t, are the maximum and minimum values ​​of the output power of the i-th type of equipment at time t respectively;

[0123] Comprehensive Demand Response Model

[0124] The IES studied in this example covers three types of load demands: electricity, gas, and heat, all of which can be managed by demand response. Among them, the electricity and gas load DR are divided into price-based DR and incentive-based DR according to different response methods. In addition, considering the characteristics of multi-energy complementary conversion, the electricity-gas and electricity-heat substitution DR are introduced. The three together constitute the IDR model.

[0125] PBDR

[0126] PBDR refers to energy companies guiding users to actively adjust their energy consumption behavior according to their own circumstances by adjusting energy prices. In actual situations, the change in load demand and the change in price are approximately exponential. In this embodiment, the exponential relationship model is used to construct the PBDR model:

[0127]

[0128] In the formula, i represents the load type, which here represents electricity and gas loads; is the class i load after PBDR; is the initial energy consumption of class i load; are the saturated upper limit load and basic load of class i load respectively; is the exponential elastic coefficient of type i load in period j to period q; is the change in electricity price of type i load in period j.

[0129] Since different types of loads have different corresponding characteristics of energy prices in different periods of time, the corresponding elasticity coefficients are also different. The differential elasticity coefficient matrix can be expressed as:

[0130]

[0131] In the formula, the matrix diagonal elements are the self-elasticity coefficients of the current period, and the non-diagonal elements are the cross-elasticity coefficients of the two periods, as shown in the following formulas:

[0132]

[0133] Among them, ΔP and ΔH are the relative increases of electricity quantity and electricity price respectively.

[0134] ADR

[0135] ADR means that users can use other energy sources to replace the required energy through various types of coupling devices. The conversion between energy sources meets the law of conservation of energy. The ADR model is as follows:

[0136]

[0137]

[0138] In the formula, They are the i loads after ADR, i represents the electricity, gas and heat loads; Represents the amount of replaced load; φ i,j Represents the conversion efficiency between energy i and energy j. This paper does not consider the conversion between hot gases, that is, φ g,h =φ h,g =0; Represents the initial load. φ i,j The calculation formula is as follows:

[0139]

[0140] W i , W j are the unit calorific values ​​of energy i and energy j respectively; η i , j are the utilization rates of energy i and energy j respectively.

[0141] IBDR

[0142] IBDR refers to energy companies using incentive subsidies to stimulate users to respond to system dispatch and adjust energy load. The IBDR model is as follows:

[0143]

[0144] In the formula, They are the i loads after IBDR, i represents the electricity and gas loads; is the load response quantity; is the load response cost of i; To compensate the price.

[0145] Demand response upper and lower limits

[0146]

[0147] In the formula, i represents the electricity, gas and heat loads. is the comprehensive demand response of the i-th type of load at time t (the sum of the above three response quantities), They are respectively the maximum and minimum values ​​of the comprehensive demand response of the i-th load at time t.

[0148] Energy balance constraints:

[0149]

[0150] In the formula, and are the electricity, heat and gas loads at IESt, They are the comprehensive demand response quantities of electricity, heat and gas loads at time t respectively. are the energy storage charging and discharging power at time t, are the thermal energy storage charging and discharging power at time t, are the charging and discharging power of gas energy storage at time t respectively.

[0151] Among them, the day-ahead optimization includes equipment constraints, energy balance constraints, and electricity and gas PBDR constraints for all load types; the one-hour, fifteen-minute, and five-minute rolling control optimizations within the day include some equipment constraints and energy balance constraints in the heat, electricity, and gas load types, respectively, as well as the comprehensive demand response constraints that must be determined on the corresponding time scale. The above-mentioned day-ahead and intraday objective functions and constraints are placed in the source-load composite control framework based on the dual-loop feedback RMPC, and the intraday rolling part is solved using the RMPC algorithm.

[0152] So far, the IES source-load composite control model based on double-loop feedback RMPC has been established. The double-feedback RMPC algorithm is used based on the CPLEX solver to obtain the output of each unit and IDR in the corresponding scheduling period, realizing the optimal control of IES. In order to verify the performance of the proposed method in dealing with uncertain problems, the following four schemes are set for comparison.

[0153] Solution 1: Ignore the randomness of source loads and adopt a deterministic scheduling strategy;

[0154] Solution 2: adopt the robust optimization scheduling strategy, with the uncertainties set to 0.8, 0.5, and 0.2 respectively;

[0155] Option 3: Use random optimization strategy;

[0156] Solution 4: Adopt a light robust optimization strategy.

[0157] The comparison results are shown in Table 1. It can be seen from the table that the deterministic optimization cost is the lowest. However, due to the large forecast errors of renewable energy and multi-energy loads, the IES load loss rate is high, and the resulting strategy cannot be applied in practice. When considering the uncertainty of source and load, the total operating costs of each algorithm have increased. Among them, the operating cost of the random optimization method is closest to the operating cost of the deterministic method, but its high load loss rate cannot guarantee the stable operation of the system. The total operating cost of the robust optimization method gradually increases with the increase of the uncertain parameter Γ, and the load loss rate gradually decreases. This is because as Γ increases, the uncertainty interval of the source and load increases, and the severity of the worst scenario increases. In order to resist greater uncertainty, the total operating cost increases. Correspondingly, since the required plan can meet the worst scenario, the load loss rate decreases. In contrast, this method has the lowest load loss rate and better economy.

[0158] Table 1 Test results of uncertain optimization algorithm

[0159]

[0160] In order to further verify the effectiveness of the proposed method, we use Figure 1 The IES shown in the figure is simulated and verified, and the optimization results are as follows Figure 4 and Figure 5 shown.

[0161] Depend on Figure 4-Figure 5It can be seen that the prediction accuracy of electric load, gas load or heat load decreases significantly with the decrease of time scale, and the prediction accuracy of the day-ahead interval of 1h deviates greatly from the actual value. Therefore, the error of the optimization scheduling strategy based on the day-ahead prediction results is often relatively large. In actual operation, the scheduling deviation can only be smoothed by purchasing energy from the external energy network, which generates a lot of additional costs and puts great pressure on the external power grid. Intraday rescheduling based on day-ahead scheduling and using more accurate prediction results can effectively avoid the above problems.

[0162] Figure 7-Figure 9 The day-ahead scheduling results are shown. Figure 7 and Figure 8 It can be seen that the CHP unit maintains a high output level almost all the time. This is because in this example, the gas price is low and the external wind and solar resources are abundant, so the CHP operation cost is low. Figure 7 It can be seen that even if the wind and solar power output is sufficient to meet the load demand, electricity is still purchased from the grid. This is because the electricity price is low. Purchasing electricity to drive other electric drive equipment to work can effectively reduce costs. Figure 8 and Fig. 9 It can be seen that both P2G equipment and EB equipment maintained a relatively high output level.

[0163] Figure 10-12 The results of intraday readjustment are shown. The load curves in the figure are the results of the intraday rolling forecast results and the demand response amount of the previous stage. In the intraday stage, the proposed method re-dispatches the thermal equipment, gas equipment and electrical equipment through three sub-controllers of different time scales, which greatly improves the flexibility of IES energy supply measurement and is of great significance for avoiding uncertain influences. In addition, the use of differentiated scheduling cycles for equipment not only effectively avoids the slow dynamic equipment (heat and gas) from being in a dynamic process at all times due to too frequent instructions, affecting the energy supply quality, but also fully stimulates the flexibility of power equipment and greatly exerts its ability to smooth out uncertainty. Fig.13 Taking the adjustment of gas equipment within 15 minutes of the day as an example, it can be found that when the source-load forecast results change within the day, the output of the equipment changes. This refined management of equipment is conducive to promoting the consumption of renewable energy.

[0164] Figure 14-17 The demand response amount at different time scales is shown. This method fully considers the differences in the response period and dynamic characteristics of different demand responses, and performs different types of demand responses at different time scales, which greatly improves the flexibility of the demand side. Figure 11-13 ,This method can fully activate the flexibility of supply-side equipment and the flexibility of demand ,response on the demand side, which is of great significance for resisting uncertainty.

[0165] Embodiment 2

[0166] This embodiment provides an IES compound control system based on dual-feedback RMPC, including:

[0167] The day-ahead source-load forecast data acquisition module is configured to: acquire day-ahead source-load forecast data of the integrated energy system;

[0168] The composite control module is configured as follows:

[0169] Based on the acquired day-ahead source-load forecast data of the integrated energy system, the electric load incentive demand response and the output of related power equipment are obtained through two stages: day-ahead dispatch and intraday correction.

[0170] Among them, in the day-ahead dispatch stage, the initial reference dispatch plan for the price-based demand response of electrical loads and electricity, heat and gas related equipment is determined;

[0171] In the intra-day correction stage, based on the price-based demand response of electrical loads and the initial reference scheduling plan for electricity, heat and gas related equipment, the intra-day correction is divided into fast, intermediate and slow control sub-layers according to the differences in the dynamic characteristics of multi-heterogeneous energy sources. Controllers with different time scales are used for differentiated control. The control time interval of each layer is used as the corresponding scale. A double-layer feedback architecture is constructed at each layer. The output disturbance of the equipment and demand response obtained in the inner loop is fed back to the outer loop. In the outer loop, the source and load data are updated, the system state prediction model is dynamically trained, and the robust optimization algorithm is guided to adjust the uncertain parameters of the system. The reference scheduling plan output by the previous layer is adjusted layer by layer, and the output of electrical load incentive demand response and related power equipment is output in the last layer.

[0172] Embodiment 3

[0173] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps in the IES multi-time scale composite control method based on dual-feedback RMPC as described above are implemented.

[0174] Embodiment 4

[0175] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the IES multi-time scale composite control method based on dual-feedback RMPC as described above are implemented.

[0176] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0177] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0178] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0179] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0180] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0181] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. The IES multi-time scale composite control method based on dual feedback RMPC is characterized by: include: Obtain day-ahead source-load forecast data for integrated energy systems; Based on the acquired day-ahead source-load forecast data of the integrated energy system, the electric load incentive demand response and the output of related power equipment are obtained through two stages: day-ahead dispatch and intraday correction. Among them, in the day-ahead dispatch stage, the initial reference dispatch plan for the price-based demand response of electrical loads and electricity, heat and gas related equipment is determined; In the intra-day correction stage, based on the price-based demand response of electrical loads and the initial reference scheduling plan for electricity, heat and gas related equipment, the intra-day correction is divided into fast, intermediate and slow control sub-layers according to the differences in the dynamic characteristics of multi-heterogeneous energy sources. Controllers with different time scales are used for differentiated control. The control time interval of each layer is used as the corresponding scale. A two-layer feedback architecture is constructed at each layer. The output disturbance of the equipment and demand response obtained by the inner loop is fed back to the outer loop. In the outer loop, the source and load data are updated, the system state prediction model is dynamically trained, and the robust optimization algorithm is guided to adjust the uncertain parameters of the system. The reference scheduling plan output by the previous layer is adjusted layer by layer, and the output of the electrical load incentive demand response and related power equipment is output at the last layer. According to the differences in dynamic characteristics of multiple heterogeneous energy sources, the intraday correction is divided into fast, intermediate and slow control sublayers, and controllers with different time scales are used for differentiated control, including: in the intraday correction stage, in the fast control sublayer, the power subsystem is taken as the control target, the control time interval is the first scale, and the first type of electricity and gas reference scheduling plan is obtained according to the latest source and load prediction data within the first scale, and is input into the intermediate control sublayer; in the intermediate control sublayer, the natural gas subsystem is taken as the control target, the control time interval is the second scale, and the second type of electricity reference scheduling plan is obtained according to the latest source and load prediction data within the second scale, and is input into the slow control sublayer; in the slow control sublayer, the power subsystem is taken as the control target, the control time interval is the third scale, and the electric load incentive demand response and related power equipment output are obtained according to the latest source and load prediction data within the third scale.

2. The IES multi-time scale composite control method based on dual feedback RMPC as claimed in claim 1, characterized in that: The first scale > the second scale > the third scale, and the three scales are coupled.

3. The IES multi-time scale composite control method based on dual feedback RMPC as claimed in claim 1, characterized in that: In the outer loop, source load data is updated, the system state prediction model is dynamically trained, and the robust optimization algorithm is guided to adjust the uncertain parameters of the system, including: The cardinality uncertainty set is used to characterize the output disturbance of equipment and demand response, and the output disturbance of equipment and demand response is incorporated into the state space equation as the intraday rolling scheduling constraint of the current scale to predict the state of the integrated energy system. The prediction accuracy of the state prediction model is evaluated in real time. Based on the accuracy of the prediction model at the current moment, the conservative parameters of the robust optimization algorithm at the next moment are adjusted, and the uncertain parameters are obtained based on the adjusted conservative parameters.

4. The IES multi-time scale composite control method based on dual feedback RMPC as claimed in claim 1, characterized in that: The day-ahead scheduling model is: ， ， ， ， In the formula, is the total operating cost of IES, is the scheduling period, For the i Energy purchase cost, For the i The operation and maintenance costs of this type of equipment, For the i Class demand response costs; is the power purchased by IES from the distribution grid, is the gas purchasing power of IES, for The price of electricity purchased from the distribution network at any time, for The natural gas price at the time, For the i Class load variation, For the system i Compensation costs for class load changes.

5. The IES multi-time scale composite control method based on dual feedback RMPC as claimed in claim 1, characterized in that: The scheduling model for the intraday correction phase is: The objective functions of the power subsystem, natural gas subsystem and power subsystem are: ， In the formula, For the i The scheduling cost of the class subsystem, For the i Equipment day t The effort of the moment, For the i Equipment day t -1 moment of output, For the i The output adjustment cost of this type of equipment, For the i Class Equipment t Always refer to the scheduling plan. For the i Output change penalty for such equipment; Constraints include equipment output constraints, comprehensive demand response constraints, and energy balance constraints; Among them, equipment output constraints include: Energy conversion equipment group: ， In the formula, , are the electrical power and thermal power output of the CHP unit respectively; , They are the electricity and gas production efficiencies of the CHP unit; Input gas power for CHP unit; , are the output gas power and input electric power of P2G respectively, is the power-to-gas efficiency; , are the output thermal power and input electrical power of EB, is the heat generation efficiency of EB; , are the output thermal power and input gas power of GB respectively, is the heat generation efficiency of GB; Renewable energy units: ， ， ， in, for The power generated by the wind turbine at the moment, and are the cut-in wind speed and cut-out wind speed of the fan respectively, is the rated wind speed, is the rated power of the fan; for The electric power output of PV at the moment, is the rated power of the PV, and They are The light intensity at the moment and the reference light intensity under standard test conditions, and are ambient temperature and reference temperature under standard test conditions, respectively; , They are t The cooling power and heat power consumed by the absorption refrigerator at any time. is the refrigeration coefficient of the absorption chiller; Hybrid energy storage system: ， In the formula, For the i Class equipment in t The output power at the moment, , Respectively i Class equipment in t The maximum and minimum values ​​of the output power at each moment; Power upper and lower limits: In the formula, For the i Class equipment in t The output power at the moment, , Respectively i Class equipment in t The maximum and minimum values ​​of the output power at each moment; Comprehensive demand response constraints include: PBDR constraints: ， In the formula, For the post-PBDR i Class load; for i Initial energy consumption of this type of load; , They are i The saturation upper limit load and base load of the load category; for i Class load j Time period q Exponential elasticity coefficient of the time period; for i Class load j The change in electricity price during the period; ADR Constraints: ， ， In the formula, After ADR i load; represents the amount of replaced load; Representative energy i and Energy j The conversion efficiency between = =0; represents the initial load; , Energy i and Energy j Unit calorific value; , Energy i and energy j Utilization rate; IBDR constraints: ， In the formula, IBDR i load, for i Load response quantity; for i load response costs; To compensate for the price; Demand response upper and lower limits: For the i Class load t The comprehensive demand response volume at the moment, , Respectively i Class load t The maximum and minimum values ​​of the comprehensive demand response at the moment; Energy balance constraints include: ， In the formula, , and are the electricity, heat and gas loads at IESt, , , are the comprehensive demand response of electricity, heat and gas loads at time t, , are the energy storage charging and discharging power at time t, , are the thermal energy storage charging and discharging power at time t, , are the charging and discharging power of gas energy storage at time t respectively.

6. The IES multi-time scale composite control method based on dual feedback RMPC as claimed in claim 3, characterized in that: The state space equation is: ， in, is the state variable matrix of the equipment output, is the control variable matrix, Output disturbance variable matrix for the controlled object, is the disturbance variable matrix, is the output variable matrix, ， ， ， , , For battery charging efficiency, is the state of charge, is the time interval, is the discharge efficiency.

7. IES multi-time scale composite control system based on dual feedback RMPC, characterized by: include: The day-ahead source-load forecast data acquisition module is configured to: acquire day-ahead source-load forecast data of the integrated energy system; The composite control module is configured as follows: Based on the acquired day-ahead source-load forecast data of the integrated energy system, the electric load incentive demand response and the output of related power equipment are obtained through two stages: day-ahead dispatch and intraday correction. Among them, in the day-ahead dispatch stage, the initial reference dispatch plan for the price-based demand response of electrical loads and electricity, heat and gas related equipment is determined; In the intra-day correction stage, based on the price-based demand response of electrical loads and the initial reference scheduling plan for electricity, heat and gas related equipment, the intra-day correction is divided into fast, intermediate and slow control sub-layers according to the differences in the dynamic characteristics of multi-heterogeneous energy sources. Controllers with different time scales are used for differentiated control. The control time interval of each layer is used as the corresponding scale. A two-layer feedback architecture is constructed at each layer. The output disturbance of the equipment and demand response obtained by the inner loop is fed back to the outer loop. In the outer loop, the source and load data are updated, the system state prediction model is dynamically trained, and the robust optimization algorithm is guided to adjust the uncertain parameters of the system. The reference scheduling plan output by the previous layer is adjusted layer by layer, and the output of the electrical load incentive demand response and related power equipment is output at the last layer. According to the differences in dynamic characteristics of multiple heterogeneous energy sources, the intraday correction is divided into fast, intermediate and slow control sublayers, and controllers with different time scales are used for differentiated control, including: In the intra-day correction stage, in the fast control sublayer, the power subsystem is taken as the control target, and the control time interval is the first scale. According to the latest source and load forecast data within the first scale, the first type of electricity and gas reference scheduling plan is obtained, which is input into the intermediate control sublayer. In the intermediate control sublayer, the natural gas subsystem is taken as the control target, and the control time interval is the second scale. According to the latest source and load forecast data within the second scale, the second type of electricity reference scheduling plan is obtained, which is input into the slow control sublayer. In the slow control sublayer, the power subsystem is taken as the control target, and the control time interval is the third scale. According to the latest source and load forecast data within the third scale, the electric load incentive demand response and the output of related power equipment are obtained.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the IES multi-time scale composite control method based on dual-feedback RMPC as described in any one of claims 1 to 6 are implemented.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the IES multi-time scale composite control method based on dual-feedback RMPC as described in any one of claims 1-6 are implemented.

Images