Micro-grid plug-and-play flexible load optimization method based on model predictive control

Abstract

The invention discloses a micro-grid plug-and-play flexible load optimization method based on model prediction control, and the method enables the load management of a micro-grid to be more flexible and efficient through the introduction of an MPC technology, enables a user to predict the future load demands and the output of distributed power generation resources, and achieves the real-time optimization of a load. By means of the optimization, dependence on traditional energy is reduced, the utilization rate of renewable energy sources is increased, and the operation cost of the system is reduced. In addition, the invention can ensure that the load is safely connected and disconnected under the physical and operational constraints of the power distribution network, which significantly improves the stability and reliability of the system. Firstly, through load aggregation and load reduction, effective management of the flexible load in the micro-grid is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of microgrids, and in particular to a microgrid plug-and-play flexible load optimization method based on model predictive control. Background Art

[0002] In recent years, the growing demand for electricity has led to increasingly serious environmental and ecological problems. With the proposal of the "dual carbon" goal, the rapid development of green and clean energy is an effective way to build an energy-saving and environmentally friendly society. The emergence of microgrid technology provides technical support for the local consumption, reasonable configuration and safe management of green renewable energy. As a new type of power system operation and management mode, the optimization and dispatching of microgrids are of great significance for improving energy utilization efficiency and ensuring the stable operation of power systems. In microgrids, traditional dispatching models are often based on static load forecasts and power generation resources, lacking effective response measures to the volatility and uncertainty of renewable energy. At the same time, they cannot achieve real-time optimization and are difficult to adapt to the rapid changes in grid conditions and demand. With the increasing penetration of renewable energy and distributed resources, the risks of voltage instability and reverse power flow also increase. Therefore, it becomes crucial to consider a wide range of distribution network constraints to coordinate loads. Flexible loads become the key object of optimal dispatching due to their adjustability. In this context, a plug-and-play flexible load optimization method for microgrids based on model predictive control (MPC) is proposed. Through real-time prediction and optimization, it can achieve effective management of flexible loads and improve the operation efficiency and economy of microgrids. Summary of the invention

[0003] The purpose of the present invention is to solve the above problems and to design a microgrid plug-and-play flexible load optimization method based on model predictive control.

[0004] To achieve the above object, the technical solution of the present invention is that, further, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the microgrid plug-and-play flexible load optimization method comprises the following steps:

[0005] Establishing a microgrid system architecture, wherein the microgrid system architecture includes at least a transferable load model, a curtailable load model, and an energy storage battery model;

[0006] Design a plug-and-play interface for the flexible load, so that the interface is connected to the microgrid;

[0007] Based on the microgrid system architecture, a predictive control algorithm is designed, and the algorithm is used to generate an operation strategy for flexible loads by predicting the system state in the future;

[0008] Determining the goal of flexible load optimization, the goal at least includes operating cost and energy utilization efficiency;

[0009] Monitor the microgrid system in real time, obtain system status information, and control the flexible load in real time based on the monitoring results and model predictive control algorithm.

[0010] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the microgrid system architecture includes:

[0011] Transferable load model,The transferable load in the microgrid system can be plug-and-play operated at different times. After connection, the load power remains unchanged, and only the connection or disconnection time changes. The equation is:

[0012] E def (t+1)=E def (t)+ΔTαP def (t)

[0013] P def (t) = α def P(t)

[0014] Where: t is the discrete time step; E def (t) represents the energy of the load that can be transferred at time t, kW·h; ΔT is the load time interval, h; α is the energy conversion loss coefficient, 0#α1; P def (t) is the power of the transferable load at time t, kW; P(t) is the total load power at time t, kW; α def is the transfer coefficient.

[0015] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the microgrid system architecture further includes:

[0016] Reducible load model: Reducible load is the load that can withstand a certain degree of power outage or power reduction, and can reduce power consumption through appropriate adjustment measures. The equation is

[0017] E shp (t+1)=E shp (t)+ΔTλP shp (t)

[0018] E shp，min (t)#E shp (t)E shp，max (t)

[0019] 0#P shp (t)P shp，max (t)

[0020] Where: E shp(t) is the energy that can be reduced at time t, kW·h; λ is the load reduction efficiency parameter, 0#λ1; P shp (t) is the power that can be reduced at time t, kW; E shp，min (t) and E shp，max (t) are the minimum and maximum values ​​of load energy that can be reduced at time t, kW·h; P shp，max (t) is the maximum power that can be reduced at time t.

[0021] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the microgrid system architecture further includes:

[0022] Energy storage battery model: The energy storage state of charge (SOC) equation and the energy storage battery linear state space model are:

[0023]

[0024] E bat (t+1)=E bat (t)+ΔT·P bat (t)

[0025] In the formula: represents the current state of charge; is the current energy storage capacity, kW; is the rated capacity of energy storage, kW; is the discharge current of energy storage at the moment, A; is the energy size of the energy storage battery at the moment, kW·h; is the power size of the energy storage battery at the moment, divided into charging and discharging power, kW; is the energy size of the energy storage battery at the moment, kW·h; the relationship between the state of charge at the previous and next moments is

[0026] S b (t) = S b (t-Δt)+η ch p ch Δt / C BE

[0027] S b (t) = S b (t-Δt)-p disc Δt / (η disc C BE )

[0028] Where: S b (t-Δt) represents the state of charge at the previous moment; η ch With η disc Represents the charging and discharging efficiency of the energy storage system; P ch With P disc They are respectively the charging and discharging power of the energy storage system, kW.

[0029] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the microgrid system architecture further includes:

[0030] Microgrid model, using the second-order cone programming method to relax the non-convex power flow model, established a Distflow power flow model:

[0031]

[0032] Where: P ij is the active power from bus i to j, kW; Q ij is the reactive power from bus i to j, kVar; p l,j 、p bat,j 、p def,j and P shp,j are the total power consumption of fixed load, battery bank, transferable load and curtailable load on line j, kW·h; q l,j and q g,j are the reactive power consumption and generation on line j, kVar; v i is the voltage amplitude on bus i, kV; L ij The square amplitude of the complex current from bus i to j, A 2 ; R ij and X ij are the resistance and reactance of line i to j, Ω;

[0033] Using the second-order cone relaxation The relaxation is as follows

[0034]

[0035] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the flexible load optimization further includes:

[0036] Optimization method of plug-and-play flexible load in microgrid based on MPC

[0037] The whole system is described as follows:

[0038] x(k+1)=Ax(k)+Bu(k)

[0039] x = [x1, x2] T =[(E shp，1 , …E shp，n ), (E bat，1 , …E bat，n )] T

[0040] u=[q g , Pshp , p bat ] T ,

[0041] Where: x(k) and x(k+1) are the state vectors of the system at time k and k+1 respectively; u(k) is the control input vector of the system at time k; E shp,n The energy required to reduce the load for the nth node; E bat,n is the energy size of the energy storage battery of the nth node; q g represents the size limit of the capacitor control input; A is the system matrix representing the evolution law of the system state; B is the input matrix representing the influence of the control input on the system state;

[0042] Define an objective function, which reflects the evaluation criteria of system performance and is usually constructed based on the difference between the actual performance of system state, control input and output and the preset target;

[0043] Set up a multiple-input multiple-output discrete-time linear system:

[0044]

[0045] Where A, B, C, and D are system matrices of appropriate sizes, and are random quantities with zero mean and certain variance, v(k) and w(k) are process noise and measurement noise respectively. u(k) and y(k) are the control input and output of the multi-input multi-output discrete-time linear system respectively.

[0046] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the multi-input multi-output discrete-time linear system further includes:

[0047] Design the corresponding optimization controller, and the objective function is designed as follows

[0048]

[0049] Where x k is the predicted state of the system at the current moment, e k is the error value of the current state, u k-1 is the control input value of the system at time k-1, N p is the prediction step size that needs to be set, and R, Q, and P are positive definite quadratic matrices of appropriate dimensions.

[0050] Furthermore, in the above-mentioned microgrid plug-and-play flexible load optimization method based on model predictive control, the multi-input multi-output discrete-time linear system further includes:

[0051] The additional deferrable load on node j modifies the set £k through the equality relation;

[0052]

[0053] represents a feasible set;

[0054] When the deferrable load requires insertion at node j and delay for d time steps, the following constraints arise:

[0055]

[0056] Where: For every possible time delay 0≤d≤d max the corresponding shift vector;

[0057] The solution of mixed integer program (x * ,u * ,z * ) gives the optimal transition time

[0058] st x(l+1)=Ax(l)+Bu(l),x0=x(0)

[0059]

[0060]

[0061] During the entire system operation, the bus voltage fluctuation range is within a small range of the nominal value;

[0062] V min #V i V max

[0063] Where: V i is the nominal value of bus voltage, KV; V min and V max They are the lower and upper limits of the bus voltage respectively.

[0064] The beneficial effect is that, through the introduction of MPC technology, the load management of the microgrid becomes more flexible and efficient, enabling users to predict future load demand and the output of distributed generation resources, thereby realizing real-time optimization of the load. This optimization not only reduces dependence on traditional energy, improves the utilization rate of renewable energy, but also reduces the operating cost of the system. In addition, the present invention can ensure the safe connection and disconnection of loads under the physical and operational constraints of the distribution network, which significantly improves the stability and reliability of the system. First, through load aggregation and load reduction, the effective management of flexible loads in the microgrid is realized, which not only improves the utilization efficiency of energy, but also reduces the risk of grid instability caused by load fluctuations; secondly, the introduction of the plug-and-play request processing mechanism enables the microgrid to quickly respond to the access and exit of distributed resources, which provides users with greater flexibility and convenience; in addition, the realization of recursive feasibility ensures the stability of the algorithm in long-term operation, which is crucial for the continuous operation of the microgrid; finally, the present invention reduces the negative impact on the environment by optimizing energy distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.

[0066] Figure 1 It is a schematic diagram of a first embodiment of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0067] Figure 2 A microgrid structure diagram containing plug-and-play flexible loads in a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0068] Figure 3 It is an MPC feedback control block diagram of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0069] Figure 4 It is a microgrid plug-and-play device execution process diagram of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0070] Figure 5 The invention discloses a distribution network plug-and-play flexible load distribution diagram of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0072] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0073] The present invention will be described in detail below in conjunction with the accompanying drawings. Figure 1 As shown, a microgrid plug-and-play flexible load optimization method based on model predictive control, the microgrid plug-and-play flexible load optimization method comprises the following steps:

[0074] Step 101: Establish a microgrid system architecture, wherein the microgrid system architecture at least includes a transferable load model, a curtailable load model, and an energy storage battery model;

[0075] Specifically, the microgrid system architecture in this embodiment includes

[0076] Transferable load model,The transferable load in the microgrid system can be plug-and-play operated at different times. After connection, the load power remains unchanged, and only the connection or disconnection time changes. The equation is:

[0077] E def (t+1)=E def (t)+ΔTαP def (t)

[0078] P def (t) = α def P(t)

[0079] Where: t is the discrete time step; E def (t) represents the energy of the load that can be transferred at time t, kW·h; ΔT is the load time interval, h; α is the energy conversion loss coefficient, 0#α1; P def (t) is the power of the transferable load at time t, kW; P(t) is the total load power at time t, kW; α def is the transfer coefficient.

[0080] Reducible load model: Reducible load is the load that can withstand a certain degree of power outage or power reduction, and can reduce power consumption through appropriate adjustment measures. The equation is

[0081] E shp (t+1)=E shp (t)+ΔTλP shp (t)

[0082] E shp，min (t)#E shp (t)E shp，max (t)

[0083] 0#P shp (t)P shp，max (t)

[0084] Where: E shp (t) is the energy that can be reduced at time t, kW·h; λ is the load reduction efficiency parameter, 0#λ1; P shp (t) is the power that can be reduced at time t, kW; E shp，min (t) and E shp，max (t) are the minimum and maximum values ​​of load energy that can be reduced at time t, kW·h; P shp，max (t) is the maximum power that can be reduced at time t.

[0085] Energy storage battery model: The energy storage state of charge (SOC) equation and the energy storage battery linear state space model are:

[0086]

[0087] E bat (t+1)=E bat (t)+ΔT·P bat (t)

[0088] In the formula: represents the current state of charge; is the current energy storage capacity, kW; is the rated capacity of energy storage, kW; is the discharge current of energy storage at the moment, A; is the energy size of the energy storage battery at the moment, kW·h; is the power size of the energy storage battery at the moment, divided into charging and discharging power, kW; is the energy size of the energy storage battery at the moment, kW·h; the relationship between the state of charge at the previous and next moments is

[0089] S b (t) = S b (t-Δt)+η ch p ch Δt / C BE

[0090] S b (t) = S b (t-Δt)-p disc Δt / (η disc C BE )

[0091] Where: S b (t-Δt) represents the state of charge at the previous moment; η ch With η disc Represents the charging and discharging efficiency of the energy storage system; P ch With P disc They are respectively the charging and discharging power of the energy storage system, kW.

[0092] Microgrid model, using the second-order cone programming method to relax the non-convex power flow model, established a Distflow power flow model:

[0093]

[0094]

[0095] Where: P ij is the active power from bus i to j, kW; Q ij is the reactive power from bus i to j, kVar; p l,j 、p bat,j 、p def,j and P shp,j are the total power consumption of fixed load, battery bank, transferable load and curtailable load on line j, kW·h; q l,j and q g,j are the reactive power consumption and generation on line j, kVar; v i is the voltage amplitude on bus i, kV; L ij The square amplitude of the complex current from bus i to j, A 2 ; R ij and X ij are the resistance and reactance of line i to j, Ω;

[0096] Using the second-order cone relaxation The relaxation is as follows

[0097]

[0098] Step 102: design a plug-and-play interface for the flexible load so that the interface is connected to the microgrid;

[0099] Step 103: Based on the microgrid system architecture, a predictive control algorithm is designed, and the algorithm is used to generate an operation strategy for the flexible load by predicting the system state within a period of time in the future;

[0100] Step 104: determining the goal of flexible load optimization, wherein the goal at least includes operating cost and energy efficiency;

[0101] Specifically, this embodiment also includes a microgrid plug-and-play flexible load optimization method based on MPC;

[0102] The whole system is described as follows:

[0103] x(k+1)=Ax(k)+Bu(k)

[0104] x = [x1, x2] T =[(E shp，1 , …E shp，n ), (E bat，1 , …E bat，n )] T

[0105] u=[q g , P shp , p bat ] T ,

[0106] Where: x(k) and x(k+1) are the state vectors of the system at time k and k+1 respectively; u(k) is the control input vector of the system at time k; E shp,n The energy required to reduce the load for the nth node; E bat,n is the energy size of the energy storage battery of the nth node; q g represents the size limit of the capacitor control input; A is the system matrix representing the evolution law of the system state; B is the input matrix representing the influence of the control input on the system state;

[0107] Define an objective function, which reflects the evaluation criteria of system performance and is usually constructed based on the difference between the actual performance of system state, control input and output and the preset target;

[0108] Set up a multiple-input multiple-output discrete-time linear system:

[0109]

[0110] Where A, B, C, and D are system matrices of appropriate sizes, and are random quantities with zero mean and certain variance, v(k) and w(k) are process noise and measurement noise respectively. u(k) and y(k) are the control input and output of the multi-input multi-output discrete-time linear system respectively.

[0111] Design the corresponding optimization controller, and the objective function is designed as follows

[0112]

[0113] Where xk is the predicted state of the system at the current moment, ek is the error value of the current state, u k-1 is the control input value of the system at time k-1, N p is the prediction step size that needs to be set, and R, Q, and P are positive definite quadratic matrices of appropriate dimensions.

[0114] The additional deferrable load on node j modifies the set £k through the equality relation;

[0115]

[0116] represents a feasible set;

[0117] When the deferrable load requires insertion at node j and delay for d time steps, the following constraints arise:

[0118]

[0119] Where: For every possible time delay 0≤d≤d max the corresponding shift vector;

[0120] The solution of mixed integer program (x * ,u * ,z * ) gives the optimal transition time

[0121] st x(l+1)=Ax(l)+Bu(l),x0=x(0)

[0122]

[0123] During the entire system operation, the bus voltage fluctuation range is within a small range of the nominal value;

[0124] V min #V i V max

[0125] Where: V i is the nominal value of bus voltage, KV; V min and V max They are the lower and upper limits of the bus voltage respectively.

[0126] Step 105: Monitor the microgrid system in real time, obtain system status information, and control the flexible load in real time based on the monitoring results and the model predictive control algorithm.

[0127] Specifically, this embodiment also includes

[0128] Build a microgrid system model

[0129] Consider the output power model of the wind power generation device affected by wind speed.

[0130] The model of the electric vehicle charging station includes charging power demand, charging time, etc.

[0131] Plug and play interface design

[0132] The electric vehicle charging pile is connected to the microgrid control system through wireless communication, achieving plug-and-play.

[0133] The control system can perform intelligent scheduling based on the demand of charging piles and the status of the microgrid.

[0134] Model Predictive Control Algorithm Design

[0135] Predict wind speed, load demand and energy storage battery status over a period of time in the future.

[0136] Comprehensively consider the output of photovoltaic power generation, wind power generation and energy storage batteries, and optimize the charging strategy of electric vehicle charging piles, such as selecting charging time and adjusting charging power.

[0137] Optimizing goal setting

[0138] The goal is to maximize energy efficiency while taking into account reducing carbon emissions and improving system reliability.

[0139] Real-time monitoring and control

[0140] Real-time monitoring of the status information of the microgrid system, including wind power generation, photovoltaic power generation, load demand, etc.

[0141] Based on the monitoring results and model predictive control algorithm, the electric vehicle charging piles are controlled in real time to achieve the optimization goals.

[0142] The beneficial effect is that, through the introduction of MPC technology, the load management of the microgrid becomes more flexible and efficient, enabling users to predict future load demand and the output of distributed generation resources, thereby realizing real-time optimization of the load. This optimization not only reduces dependence on traditional energy, improves the utilization rate of renewable energy, but also reduces the operating cost of the system. In addition, the present invention can ensure the safe connection and disconnection of loads under the physical and operational constraints of the distribution network, which significantly improves the stability and reliability of the system. First, through load aggregation and load reduction, the effective management of flexible loads in the microgrid is realized, which not only improves the utilization efficiency of energy, but also reduces the risk of grid instability caused by load fluctuations; secondly, the introduction of the plug-and-play request processing mechanism enables the microgrid to quickly respond to the access and exit of distributed resources, which provides users with greater flexibility and convenience; in addition, the realization of recursive feasibility ensures the stability of the algorithm in long-term operation, which is crucial for the continuous operation of the microgrid; finally, the present invention reduces the negative impact on the environment by optimizing energy distribution.

[0143] Specifically, the technical solution of the present invention also includes the following methods:

[0144] Microgrid system structure

[0145] The grid-connected microgrid consisting of photovoltaic, wind power, micro-turbines, fuel cells, batteries and loads is selected as the research object. The distributed power supply, energy storage and loads in the microgrid are all single controllable units. Figure 1 As shown in the figure, when the distributed power output cannot meet the load demand, it will be compensated by the grid, and when the distributed power output is surplus, it can transmit electricity to the grid; multiple microgrids are connected to the power line. This two-way energy flow mechanism not only improves the efficiency of energy utilization, but also helps to reduce the impact on the environment. This grid-connected microgrid achieves efficient energy utilization and stable operation of the system by integrating multiple energy forms and energy storage equipment, providing an effective solution for the sustainable development of the power system.

[0146] System key equipment model

[0147] This section will introduce the various modules in the power line, including shiftable loads, curtailable loads, battery banks, and microgrid models.

[0148] ① Transferable load model: The transferable load in the microgrid system can be plug-and-play operated at different times. After connection, the load power remains unchanged, and only the connection or disconnection time changes. The equation is

[0149] E def (t+1)=E def (t)+ΔTαP def (t) (1)

[0150] P def (t) = α def P(t) (2)

[0151] Where: t is the discrete time step; E def (t) represents the energy of the load that can be transferred at time t, kW·h; ΔT is the load time interval, h; α is the energy conversion loss coefficient, 0#α1; P def (t) is the power of the transferable load at time t, kW; P(t) is the total load power at time t, kW; α def is the transfer coefficient.

[0152] ② Reducible load model: Reducible load is the load that can withstand a certain degree of power outage or power reduction, and can reduce power consumption through appropriate adjustment measures. The equation is

[0153] E shp (t+1)=E shp (t)+ΔTλP shp (t) (3)

[0154] E shp，min (t)#E shp (t)E shp，max (t) (4)

[0155] 0#P shp (t)P shp，max (t) (5)

[0156] Where: E shp (t) is the energy that can be reduced at time t, kW·h; λ is the load reduction efficiency parameter, 0#λ1; P shp (t) is the power that can be reduced at time t, kW; E shp，min (t) and E shp，max (t) are the minimum and maximum values ​​of load energy that can be reduced at time t, kW·h; P shp，max (t) is the maximum power of the load that can be reduced at time t;

[0157] ③Energy storage battery model: The energy storage state of charge (SOC) equation and the energy storage battery linear state space model are:

[0158]

[0159] E bat (t+1)=E bat (t)+ΔT·P bat (t) (7)

[0160] Where: S b (t) represents the current state of charge; C NET is the current energy storage capacity, kW; C BS is the rated capacity of energy storage, kW; I t is the discharge current of energy storage at time t, A; E bat (t) is the energy of the energy storage battery at time t, kW·h; P bat (t) is the power of the energy storage battery at time t, which is divided into charging and discharging power, kW; E bat (t) is the energy of the energy storage battery at time t, kW·h; and the relationship between the state of charge at the previous and next two moments is:

[0161] S b (t) = S b (t-Δt)+η ch p ch Δt / C BE (8)

[0162] S b (t) = S b (t-Δt)-p disc Δt / (η disc C BE ) (9)

[0163] Where: S b (t-Δt) represents the state of charge at the previous moment; η ch With η disc Represents the charging and discharging efficiency of the energy storage system; P ch With P disc They are respectively the charging and discharging power of the energy storage system, kW.

[0164] ④Microgrid model: The second-order cone programming method is used to relax the non-convex power flow model, and a Distflow-based power flow model is established.

[0165]

[0166] Where: P ij is the active power from bus i to j, kW; Q ij is the reactive power from bus i to j, kVar; p l,j 、p bat,j 、p def,j and P shp,j are the total power consumption of fixed load, battery bank, transferable load and curtailable load on line j, kW·h; q l,j and q g,j are the reactive power consumption and generation on line j, kVar; v iis the voltage amplitude on bus i, kV; L ij The square amplitude of the complex current from bus i to j, A 2 ; R ij and X ij are the resistance and reactance of line i to j, Ω;

[0167] Using the second-order cone relaxation, equation (13) is relaxed as follows:

[0168]

[0169] Optimization method of plug-and-play flexible load in microgrid based on MPC

[0170] The whole system is described as follows:

[0171]

[0172] Where: x(k) and x(k+1) are the state vectors of the system at time k and k+1 respectively; u(k) is the control input vector of the system at time k; E shp,n The energy required to reduce the load for the nth node; E bat,n is the energy size of the energy storage battery of the nth node; q g represents the size limit of the capacitor control input; A is the system matrix representing the evolution law of the system state; B is the input matrix representing the influence of the control input on the system state.

[0173] Model predictive control (MPC) is an advanced optimization-based control strategy that has been widely used in many fields such as industrial automation, process control, and stability management of microgrids. Compared with traditional control methods, MPC uses a mathematical model of the system to predict future behavior. It combines real-time measurement data with preset target data and calculates control inputs through an optimization process. This advanced control method can ensure that the system continues to operate stably in complex and dynamic environments, making MPC an ideal choice for handling variable, linear, and nonlinear systems. The optimization control problem can be further transformed into a mathematical optimization problem consisting of an objective function and constraints. When constructing a model predictive control (MPC) algorithm, it is first necessary to define an objective function, which reflects the evaluation criteria for system performance and is usually constructed based on the difference between the actual performance of the system state, control inputs, and outputs and the preset targets. Then, in order to ensure that the system can comply with physical laws and operating constraints in actual operation, these constraints need to be incorporated into the model in mathematical form. Consider the following multi-input and multi-output discrete-time linear system

[0174]

[0175] Where A, B, C, and D are system matrices of appropriate sizes, and are random quantities with zero mean and certain variance, v(k) and w(k) are process noise and measurement noise respectively. u(k) and y(k) are the control input and output of system (18) respectively. For the above system (18), the corresponding optimization controller is designed, and the objective function is designed as follows

[0176]

[0177] Where x k is the predicted state of the system at the current moment, e k is the error value of the current state, u k-1 is the control input value of the system at time k-1, N p is the prediction step size to be set, R, Q, and P are positive definite quadratic matrices of appropriate dimensions. Equation (19) takes into account the system state-related targets, error targets, and input targets. The MPC feedback control block diagram is shown in the attached figure. Figure 3 .

[0178] In the actual power distribution system, users can connect or disconnect their appliances at will. The execution process of the microgrid plug-and-play device is as shown in the attached figure. Figure 4 This changes the overall load on the system and has a significant impact on the bus voltage. The MPC scheme is extended to the case where the system dynamics in (15) change due to loads joining or leaving the network.

[0179] As shown in formula (5), the curtailable load can be inserted without drawing energy from the grid: 0≤p shp ≤p shp,max Therefore, for the curtailable load, as long as it applies, it should be inserted immediately, and it can be shp = 0, and wait for the system to allow strictly positive values ​​of p shp >0.

[0180] The goal of the plug-and-play operation is to find the time to safely connect a deferrable load and to modify the responses of the deformable load and control devices so that the connection is achieved as quickly as possible. This section defines a mixed integer procedure for finding the minimum time to safely connect a deferrable load. Deferrable loads do not affect the dynamics of the system, they only change the feasible set. An additional deferrable load at node j modifies the set ￡k by the equality relation:

[0181]

[0182] Additionally, it modifies the terminal set In the following, (respectively ) denotes the feasible set (respectively the terminal set), and these constraints remain unchanged when the deferrable load is inserted. The deferrable load makes a plug-and-play request at time k, which results in d max possible load shapes.

[0183] Therefore, when the deferrable load requires insertion at node j and delay for d time steps, the following constraints arise:

[0184]

[0185]

[0186] Where: For every possible time delay 0≤d≤d max The corresponding shift vector.

[0187] The solution of mixed integer program (x * ,u * ,z * ) gives the optimal transition time

[0188] st x(l+1)=Ax(l)+Bu(l),x0=x(0)

[0189]

[0190] Equation (25) minimizes the transition delay and ensures that the problem remains feasible when the load is inserted. Equation (26) and the mixed integer program ensure that all constraints are satisfied during the transition time and that the modified network is satisfied. The entire program remains feasible during the transition and ensures recursive feasibility after the modification.

[0191] During the operation of the entire system, in order to ensure the reliability of the distribution network operation, the bus voltage fluctuation range is required to be within a smaller range of the nominal value:

[0192] V min #V i V max 27)

[0194] Where: V i is the nominal value of bus voltage, KV; V min and V max They are the lower and upper limits of the bus voltage respectively.

[0195] In order to prevent the reverse power supply of energy storage equipment in the microgrid from causing line power flow, the active power of all buses is required to be positive:

[0196] P ij 3 0 28)

[0198] 5. Effects of the invention

[0199] Through the introduction of MPC technology, the load management of microgrids becomes more flexible and efficient, enabling users to predict future load demand and the output of distributed generation resources, thereby achieving real-time optimization of loads. This optimization not only reduces dependence on traditional energy, improves the utilization rate of renewable energy, but also reduces the operating cost of the system. In addition, the present invention can ensure the safe connection and disconnection of loads under the physical and operational constraints of the distribution network, which significantly improves the stability and reliability of the system.

[0200] Firstly, through load aggregation and load reduction, effective management of flexible loads in microgrids is achieved, which not only improves the energy utilization efficiency, but also reduces the risk of grid instability caused by load fluctuations; secondly, the introduction of plug-and-play request processing mechanism enables microgrids to quickly respond to the access and exit of distributed resources, which provides users with greater flexibility and convenience; in addition, the realization of recursive feasibility ensures the stability of the algorithm in long-term operation, which is crucial for the continuous operation of microgrids; finally, the present invention reduces the negative impact on the environment by optimizing energy distribution.

[0201] In order to verify the effectiveness of the microgrid plug-and-play flexible load optimization method based on model predictive control, the modified IEEE33 node distribution line is analyzed. Node 1 is the first node of the line. The transformer (SZ11-31500 / 110) between node 0 and node 1 simulates the equivalent impedance of the upper power grid of the line. Node 0 is used as a balancing node, and the energy storage battery is at nodes 2, 7, 14, and 21. The voltage level is 10kV, the minimum power factor allowed for the load and line is 0.95, and the maximum total load of the network is 3.715+j1.2122 MV·A. The active power loss is about 4%, the correction coefficient a is 1.04, the error of the load prediction curve is 10%, and the power flow reverse coefficient γ is 0.3.

[0202] Nodes 2, 10, 15 and 18 are used as DG access points, the allowable range of voltage deviation is ±5%, the range of DG power factor is 0.95 (leading) to 0.95 (lagging), and the fluctuation curves of load and DG are as follows: Figure 5 shown.

[0203] Table 1 Plug and Play Operation of Delayable Loads

[0204]

[0205] The results are attached Figure 5, a model predictive control method that minimizes the total system cost is used to aggregate flexible loads and provide load shaping targets under distribution network constraints. Secondly, a mixed integer programming (MIP) is defined to safely connect loads and minimize waiting time. It is proved that the algorithm achieves recursive feasibility by appropriately defining the connection conditions and terminal constraint sets; and improves the operation efficiency of microgrids, reduces operating costs, and enhances the integration capability of renewable energy, thereby promoting the sustainable development of power systems.

[0206] (1) Application of Model Predictive Control (MPC): This patent uses MPC technology to optimize the flexible load in the microgrid and achieves real-time adjustment and optimization of the load by predicting the future system status and load demand.

[0207] (2) Load aggregation and load reduction: Through the MPC method, the dispersed flexible loads are aggregated, and a load shaping strategy is formulated according to the operating status of the system and the constraints of the power grid.

[0208] (3) Plug and Play (P&P) request processing: The patent proposes a mechanism that can handle plug and play requests of distributed resources such as electric vehicles and energy storage devices to ensure the stable operation of the microgrid.

[0209] (4) Mixed Integer Programming (MIP): In order to safely connect and disconnect loads under the constraints of the distribution network, the patent defines MIP to minimize waiting time and ensure the safety of operation.

[0210] (5) Recursive feasibility: The patent proves that the proposed algorithm can achieve recursive feasibility by properly defining the connection conditions and terminal constraint sets, ensuring long-term stable operation.

[0211] By protecting these key points and innovations, the patent aims to ensure the invention’s proprietary technology and market advantages while providing an effective solution for load management and optimization of microgrids.

[0212] See also Figure 2 , a microgrid structure diagram containing plug-and-play flexible loads of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0213] See also Figure 3 , an MPC feedback control block diagram of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0214] See also Figure 4 , a microgrid plug-and-play device execution process diagram of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention;

[0215] See also Figure 5 , a distribution network plug-and-play flexible load distribution diagram of a microgrid plug-and-play flexible load optimization method based on model predictive control in an embodiment of the present invention.

[0216] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A microgrid plug-and-play flexible load optimization method based on model predictive control, characterized in that: The microgrid plug-and-play flexible load optimization method comprises the following steps: Establishing a microgrid system architecture, wherein the microgrid system architecture includes at least a transferable load model, a curtailable load model, and an energy storage battery model; Design a plug-and-play interface for the flexible load, so that the interface is connected to the microgrid; Based on the microgrid system architecture, a predictive control algorithm is designed, and the algorithm is used to generate an operation strategy for flexible loads by predicting the system state in the future; Determining the goal of flexible load optimization, the goal at least includes operating cost and energy utilization efficiency; Monitor the microgrid system in real time, obtain system status information, and control the flexible load in real time based on the monitoring results and model predictive control algorithm.

2. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 1, characterized in that: The microgrid system architecture includes: Transferable load model,The transferable load in the microgrid system can be plug-and-play operated at different times. After connection, the load power remains unchanged, and only the connection or disconnection time changes. The equation is: E def (t+1)=E def (t)+ΔTαP def (t) P def (t)=α def P(t) Where: t is the discrete time step; E def (t) represents the energy of the load that can be transferred at time t, kW·h; ΔT is the load time interval, h; α is the energy conversion loss coefficient, 0#α1; P def (t) is the power of the transferable load at time t, kW; P(t) is the total load power at time t, kW; α def is the transfer coefficient.

3. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 1, characterized in that: The microgrid system architecture further includes: Reducible load model: Reducible load is the load that can withstand a certain degree of power outage or power reduction and can reduce power consumption through appropriate adjustment measures. The equation is: E shp (t+1)=E shp (t)+ΔTλP shp (t) E shp，min (t)#E shp (t)E shp，max (t) 0#P shp (t)P shp，max (t) Where: E shp (t) is the energy that can be reduced at time t, kW·h; λ is the load reduction efficiency parameter, 0#λ1; P shp (t) is the power that can be reduced at time t, kW; E shp，min (t) and E shp，max (t) are the minimum and maximum values ​​of load energy that can be reduced at time t, kW·h; P shp，max (t) is the maximum power that can be reduced at time t.

4. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 1, characterized in that: The microgrid system architecture further includes: Energy storage battery model: The energy storage state of charge (SOC) equation and the energy storage battery linear state space model are: E bat (t+1)=E bat (t)+ΔT·P bat (t) Where: S b (t) represents the current state of charge; C NET is the current energy storage capacity, kW; C BS is the rated capacity of energy storage, kW; I t is the discharge current of energy storage at time t, A; E bat (t) is the energy of the energy storage battery at time t, kW·h; P bat (t) is the power of the energy storage battery at time t, which is divided into charging and discharging power, kW; E bat (t) is the energy of the energy storage battery at time t, kW·h; the relationship between the state of charge at the previous and next two moments is: S b (t)=S b (t-Δt)+η ch p ch Δt / C BE S b (t)=S b (t-Δt)-p disc Δt / (η disc C BE ) Where: S b (t-Δt) represents the state of charge at the previous moment; η ch With η disc Represents the charging and discharging efficiency of the energy storage system; P ch With P disc They are respectively the charging and discharging power of the energy storage system, kW.

5. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 1, characterized in that: The microgrid system architecture further includes: Microgrid model, using the second-order cone programming method to relax the non-convex power flow model, established a Distflow power flow model: Where: P ij is the active power from bus i to j, kW; Q ij is the reactive power from bus i to j, kVar; p l,j 、p bat,j 、p def,j and P shp,j are the total power consumption of fixed load, battery bank, transferable load and curtailable load on line j, kW·h; q l,j and q g,j are the reactive power consumption and generation on line j, kVar; v i is the voltage amplitude on bus i, kV; L ij The square amplitude of the complex current from bus i to j, A 2 ; R ij and X ij are the resistance and reactance of line i to j, Ω; Using the second-order cone relaxation The relaxation is as follows:

6. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 1, characterized in that: The flexible load optimization further includes: Microgrid plug-and-play flexible load optimization method based on MPC; The whole system is described as follows: x(k+1)=Ax(k)+Bu(k) x=[x1,x 2 ] T =[(And shp,1 ,…AND shp，n ),(AND bat，n ,…AND bat，n )] T Where: x(k) and x(k+1) are the state vectors of the system at time k and k+1 respectively; u(k) is the control input vector of the system at time k; E shp,n The energy required to reduce the load for the nth node; E bat,n is the energy size of the energy storage battery of the nth node; q g represents the size limit of the capacitor control input; A is the system matrix representing the evolution law of the system state; B is the input matrix representing the influence of the control input on the system state; Define an objective function, which reflects the evaluation criteria of system performance and is usually constructed based on the difference between the actual performance of system state, control input and output and the preset target; Set up a multiple-input multiple-output discrete-time linear system: Where A, B, C, and D are system matrices of appropriate sizes, and are random quantities with zero mean and certain variance, v(k) and w(k) are process noise and measurement noise respectively, u(k) and y(k) are the control input and output of the multi-input and multi-output discrete-time linear system respectively.

7. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 6, characterized in that: The multi-input multi-output discrete-time linear system further comprises: Design the corresponding optimization controller, and the objective function is designed as follows: Where x k is the predicted state of the system at the current moment, e k is the error value of the current state, u k-1 is the control input value of the system at time k-1, N p is the prediction step size that needs to be set, and R, Q, and P are positive definite quadratic matrices of appropriate dimensions.

8. A microgrid plug-and-play flexible load optimization method based on model predictive control as claimed in claim 6, characterized in that: The multi-input multi-output discrete-time linear system further comprises: The additional deferrable load on node j modifies the set £k through the equality relation; represents a feasible set; When the deferrable load requires insertion at node j and delay for d time steps, the following constraints arise: Where: For every possible time delay 0≤d≤d max The corresponding shift vector; The solution of mixed integer program (x * ,u * ,z * ) gives the optimal transition time During the entire system operation, the bus voltage fluctuation range is within a small range of the nominal value; V min #V i V max Where, V i is the nominal value of bus voltage, KV; V min and V max They are the lower and upper limits of the bus voltage respectively.

Images