A multi-process-based microgrid distributed control method and system

Abstract

The application provides a kind of microgrid distributed control method and system based on multi-process, belong to alternating current microgrid distributed control field, method includes: in three-level process, with alternating direction multiplier algorithm to minimize the total reactive power loss in networked microgrid as the goal to optimize networked microgrid;In two-level process, using multi-agent consensus algorithm for distributed consistency control, in main process, adopt droop control to each microgrid, adopt droop control to the distributed power generation unit in single microgrid, then carry out secondary linear control, to realize the frequency recovery, voltage recovery of distributed power generation unit in single microgrid and the active power distribution, reactive power distribution between multiple power generation units.The application can effectively carry out voltage frequency recovery and power reasonable distribution, avoid the problem of frequency and voltage deviation.

Description

Technical Field

[0001] The present invention belongs to the field of AC microgrid distributed control, and in particular relates to a multi-process-based microgrid distributed control method and system. Background Art

[0002] In order to protect the ecological environment and achieve sustainable development, renewable energy is becoming increasingly attractive to future microgrids (MG), and people are increasingly demanding environmental friendliness, scalability, and flexibility. With the rapid development of new energy power generation, distributed generation (DG) represented by photovoltaics and wind turbines has the advantages of less pollution, high reliability, high energy efficiency, low transmission cost, and flexible installation location. However, the direct connection of a large number of distributed power sources to the microgrid will have a great impact on the voltage stability of the microgrid, and seriously affect the power flow distribution, power sharing, and relay protection of the microgrid. Therefore, how to maintain the efficient, safe, economical and stable operation of the microgrid has become a problem that must be solved for the high-quality development of the future microgrid.

[0003] Most existing research on AC microgrids focuses on single MG systems. However, in extreme situations, such as main grid failures and natural disasters, a single MG system may not be able to ensure reliable operation. To improve the reliability and resilience of the entire system, a feasible solution is to interconnect MGs within a certain area to form a networked microgrid (NMG) system. Practical application scenarios of NMG systems include active distribution networks including residential microgrids, building microgrid communities, and offshore power systems including port and shipboard microgrids. NMG systems can be further classified based on their electrical structure, such as voltage level, current type, and connection method.

[0004] In addition, the application of edge agents in distributed AC microgrids can optimize energy management, improve the system's real-time response capabilities, and help promote the on-site consumption and sustainable development of distributed energy.

[0005] Some scholars have explored the distributed control strategy of AC microgrids and proposed many different control concepts, which are mainly divided into three types: (1) active and reactive power (PQ) control; (2) voltage and frequency (V / f) control; and (3) droop control. Among them, PQ control is mainly used in the scenario of AC microgrid grid-connected operation. V / f control is an open-loop control method that aims to keep the voltage and frequency of the grid at a specified value by adjusting the output, but its control accuracy is low. Droop control is a closed-loop control strategy that adjusts the voltage and frequency of the AC microgrid by simulating the droop characteristics of traditional grid generators. It has high control accuracy and is suitable for microgrid environments with precise power distribution and coordinated control, but it will bring about frequency and voltage deviation problems. In order to ensure the flexible and stable operation of NMG, some scholars applied the single MG control framework to the NMG system and proposed a two-layer distributed control scheme, but did not study the coordination problem with the upper-layer optimization scheduling. Summary of the Invention

[0006] In response to the deficiencies of the existing technology, this application proposes a multi-process based microgrid distributed control method and system.

[0007] In a first aspect, the present application proposes a multi-process-based distributed control method for a microgrid, comprising:

[0008] Establishing an edge agent for each microgrid in a networked microgrid, each edge agent comprising: a main process, a secondary process, and a tertiary process, wherein the networked microgrid comprises a plurality of microgrids, each microgrid comprising a plurality of distributed generation units;

[0009] In the third-level process, the alternating direction multiplier algorithm is used to optimize the networked microgrid with the goal of minimizing the total reactive power loss in the networked microgrid. The active power reference value and reactive power reference value of each microgrid are obtained and sent to the second-level process.

[0010] In the secondary process, a multi-agent consensus algorithm is used for distributed consistency control based on local measurement signals, measurement signals from adjacent microgrids, and active and reactive power reference values ​​from the tertiary process. This allows for frequency recovery of the microgrid, voltage recovery at the point of common coupling, and arbitrary power sharing between microgrids. The calculated voltage and frequency control signals are then sent to the droop controller in the primary process.

[0011] In the main process, droop control is adopted for each microgrid to achieve autonomous operation. According to the voltage control signal and frequency control signal from the secondary process, the voltage reference value and frequency reference value are calculated and sent to the droop controller of the power generation unit.

[0012] The droop control is adopted for the distributed power generation units in the single micro-grid, and secondary linear control is performed to track the voltage reference value and the frequency reference value fed in the main process, so as to realize frequency recovery and voltage recovery of the distributed power generation units in the single micro-grid and active power distribution and reactive power distribution among the multiple power generation units.

[0013] The micro-grid distributed control method based on the multiple processes further comprises: setting a trigger function for communication of the adjacent micro-grid, and acquiring voltage state and frequency state of the adjacent micro-grid when a value of the trigger function meets a preset threshold, so as to update voltage state and frequency state of the local micro-grid and propagate the voltage state and the frequency state of the local micro-grid to the adjacent micro-grid.

[0014] The trigger function has the following calculation formula:

[0015]

[0016] Wherein, f i (t) is the trigger function of the ith micro-grid at the tth moment, σ i is a first setting coefficient, α i is a second setting coefficient, and 0 < σ i < 1, 0 < α i < 1 / d i , d i is the in-degree of the ith micro-grid, ζ avg,i is the mean value of the voltage or the mean value of the frequency of the ith micro-grid, e i (t) is a measurement error, and has the following calculation formula:

[0017]

[0018] Wherein, e i (t) is the measurement error of the ith micro-grid at the tth moment, is the measurement value of the ith micro-grid at the tth moment, i.e. voltage and frequency, is the measurement value of the ith micro-grid at the tth k moment, t k+1 is the tth moment, t k+1 is the tth moment. k k

[0019] The alternating direction multiplier algorithm is adopted to optimize the networked micro-grid with the control target of minimizing total reactive power loss in the networked micro-grid, so as to obtain the active power reference value and the reactive power reference value of each micro-grid, which comprises:

[0020] An optimal power flow problem is established with the control target of minimizing total reactive power loss in the networked micro-grid.​​

[0021] Decompose the optimal power flow problem into multiple sub-problems according to the number of buses;

[0022] Solve each sub-problem to obtain the active power reference value and reactive power reference value of each microgrid and send them to the secondary process.

[0023] In the secondary process, a multi-agent consensus algorithm is used to perform distributed consistency control based on local measurement signals, measurement signals from adjacent microgrids, and active power reference values ​​and reactive power reference values ​​from the tertiary process to achieve frequency recovery of the microgrid, voltage recovery at the point of common coupling, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the main process, including:

[0024] Based on the local frequency measurement signal and active power measurement signal of each microgrid, the frequency measurement signal and active power measurement signal from the adjacent microgrid, and the active power reference value from the three-level process, a multi-agent consensus algorithm is used to perform distributed consistency control on the frequency of each microgrid to obtain the frequency control signal. The calculation formula is as follows:

[0025]

[0026] in, For microgrid MG i The frequency control signal, N is the total number of microgrids, g i MG i The pinning gain, MG i stands for Local Microgrid, MG j represents the adjacent microgrid, a ij For local microgrid MG i With adjacent microgrid MG j The weighted adjacency matrix of For local microgrid MG i The active power reference value of For adjacent microgrid MG j The active power reference value of To adjust MG i The droop coefficient of active power, To adjust MG j The droop coefficient of active power, For local microgrid MG i The frequency measurement value, For adjacent microgrid MG j The frequency measurement value, For local microgrid MG i The measured active power value of For adjacent microgrid MG j The measured active power value of For the updated local microgrid MG i frequency control signal.

[0027] According to the local voltage signal measurement signal and reactive power measurement signal of each microgrid and the voltage signal measurement signal and reactive power measurement signal from the adjacent microgrid and the reactive power reference value from the three-level process, a multi-agent consensus algorithm is used to perform distributed consistency control on the voltage of each microgrid to obtain the voltage control signal. The calculation formula is as follows:

[0028]

[0029] in, For microgrid MG i The voltage control signal, N is the total number of microgrids, MG i stands for Local Microgrid, MG j represents the adjacent microgrid, a ij For local microgrid MG i With adjacent microgrid MG j The weighted adjacency matrix of For local microgrid MG i The reactive power reference value, For adjacent microgrid MG j Reactive power reference value, V * is the voltage standard value, To adjust MG i The droop factor of reactive power, To adjust MG j The droop factor of reactive power, For adjacent microgrid MG j The voltage control signal, For local microgrid MG i The measured value of reactive power, For adjacent microgrid MG j The measured value of reactive power, V PCC is the voltage at the point of common coupling, For the updated local microgrid MG i voltage control signal.

[0030] In the secondary process, a multi-agent consensus algorithm is used for distributed consistency control, and the frequency and common coupling voltage of the microgrid will converge to the standard value, which is expressed as follows:

[0031]

[0032]

[0033]

[0034]

[0035] Among them, ω * is the standard value of frequency, V * is the voltage standard value, is the local microgrid MG at time t i Frequency, V PCC is the voltage at the point of common coupling, To adjust MG j The droop coefficient of active power, To adjust MG i The droop coefficient of active power, is the adjacent microgrid MG at time t j The measured value of active power, is the local microgrid MG at time t i The measured value of active power, For adjacent microgrid MG j The active power reference value of For local microgrid MG i The active power reference value of To adjust MG j The droop factor of reactive power, To adjust MG i The droop factor of reactive power, is the adjacent microgrid MG at time t j The measured value of reactive power, is the local microgrid MG at time t i The measured value of reactive power, For local microgrid MG i The reactive power reference value, For adjacent microgrid MG j The reactive power reference value.

[0036] In the main process, droop control is adopted for each microgrid to achieve autonomous operation. The calculation formula is as follows:

[0037]

[0038]

[0039] Among them, ω * is the standard value of frequency, V * is the voltage standard value, is the frequency control signal from the secondary process; is the voltage control signal from the secondary process; To adjust MG i The droop coefficient of active power, To adjust MG i The droop factor of reactive power, For microgrid MG i The calculated frequency value is used as the frequency reference value sent to the distributed generation unit and as the frequency measurement value in the secondary process. For microgrid MG i The calculated voltage value is used as the voltage reference value sent to the distributed generation unit and as the voltage measurement value in the secondary process. For microgrid MG i The measured value of active power, For microgrid MG i A measurement of reactive power.

[0040] The distributed generation unit in a single microgrid adopts droop control, and the calculation formula is as follows:

[0041]

[0042]

[0043] in, is the frequency calculation value of the power generation unit, is the calculated value of the voltage of the power generation unit, For the microgrid MG i The frequency reference value, For the microgrid MG i The voltage reference value, To adjust the power generation unit DG k The droop coefficient of active power, To adjust the power generation unit DG k The ranges of reactive power droop coefficient are:

[0044]

[0045]

[0046] in, To control the maximum frequency of the power generation unit, To control the minimum frequency of the generating unit, DG is the power generation unit k The maximum value of active power, DG is the power generation unit k The minimum value of active power, DG is the power generation unit k The maximum value of the voltage, DG is the power generation unitk The minimum value of the voltage, DG is the power generation unit k The maximum value of reactive power, DG is the power generation unit k Minimum value of reactive power.

[0047] The aforementioned secondary linear control is performed on the distributed power generation unit, and the calculation formula is as follows:

[0048]

[0049]

[0050]

[0051]

[0052] in, For microgrid MG i Medium power generation unit DG k and power generation unit DG h Communication coefficient between the power generation unit DG k and power generation unit DG h If there is a link between otherwise For microgrid MG i Medium power generation unit DG k The restraining gain of the power generation unit DG k Can receive directly and but otherwise DG is the power generation unit k The calculated frequency value of To control the power generation unit DG h The calculated frequency value of Adjusting the power generation unit DG k The droop coefficient of active power, Adjusting the power generation unit DG h The droop coefficient of active power, DG is the power generation unit k The measured value of active power, DG is the power generation unit k The frequency control signal, For microgrid MG i The total number of power generation units in DG is the power generation unit h The measured value of active power, DG is the power generation unit k The calculated voltage value, DG is the power generation unit k The droop factor of reactive power, DG is the power generation unit k The measured value of reactive power, DG is the power generation unit k The voltage control signal, DG is the power generation unit h The voltage control signal, For the microgrid MG i The frequency reference value, For the microgrid MG i The voltage reference value, DG is the power generation unit k The calculated voltage value, DG is the power generation unit h The droop factor of reactive power, DG is the power generation unit h The measured value of reactive power, DG is the power generation unit k The measured value of reactive power, DG is the power generation unit k The droop factor of reactive power, DG is a distributed generation unit k an updated frequency control signal; DG is a distributed generation unit k Updated voltage control signal.

[0053] In a second aspect, the present application proposes a multi-process-based microgrid distributed control system, comprising: an edge agent module, a three-level process module, a two-level process module, a main process module, and a single microgrid control module, wherein the edge agent module is connected to the three-level process module, the two-level process module, and the main process module, respectively; the three-level process module is connected to the two-level process module, the two-level process module is connected to the main process module, and the main process module is connected to the single microgrid control module;

[0054] An edge agent module is used to establish an edge agent for each microgrid in the networked microgrid, each edge agent including: a main process, a secondary process and a tertiary process, the networked microgrid including multiple microgrids, each microgrid including multiple distributed power generation units;

[0055] The third-level process module is used to optimize the networked microgrid by adopting an alternating direction multiplier algorithm in the third-level process to minimize the total reactive power loss in the networked microgrid, obtain the active power reference value and the reactive power reference value of each microgrid, and send them to the second-level process;

[0056] The secondary process module is used to perform distributed consistency control in the secondary process using a multi-agent consensus algorithm based on local measurement signals, measurement signals from adjacent microgrids, and active power reference values ​​and reactive power reference values ​​from the tertiary process to achieve frequency recovery of the microgrid, voltage recovery at the common coupling point, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the main process.

[0057] The main process module is used to implement droop control for each microgrid in the main process to achieve autonomous operation. The voltage reference value and frequency reference value are calculated based on the voltage control signal and frequency control signal from the secondary process, and the voltage reference value and frequency reference value are sent to the droop controller of the power generation unit.

[0058] A single microgrid control module is used to adopt droop control on the distributed generation units in a single microgrid, and then perform quadratic linear control to track the voltage reference value and frequency reference value fed into the main process, so as to achieve frequency recovery, voltage recovery of the distributed generation units in the single microgrid, and active power distribution and reactive power distribution among multiple generation units.

[0059] Beneficial effects:

[0060] This application proposes a multi-process distributed control method and system for microgrids. This system employs a hierarchical control approach for coordinated control across multiple processes within a networked microgrid (NMG) system. This method effectively restores voltage and frequency, while also ensuring proper power distribution between microgrids. This avoids frequency and voltage deviations and addresses optimal scheduling under optimal power flow. While meeting control requirements, the design of an event-triggered mechanism reduces communication times and conserves communication resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 Schematic diagram of a multi-process distributed control process of a microgrid according to an embodiment of the present application;

[0062] Figure 2 A multi-process distributed control flow chart of a microgrid according to an embodiment of the present application;

[0063] Figure 3 A block diagram of a microgrid distributed control based on multiple processes according to an embodiment of the present application;

[0064] Figure 4 The topology diagram of the NMG system in the embodiment of the present application is a diagram taking four MG units as an example;

[0065] Figure 5 Frequency, voltage and active power simulation example diagram of the embodiment of the present application;

[0066] Figure 6 Principle block diagram of a multi-process based microgrid distributed control system according to an embodiment of the present application. DETAILED DESCRIPTION

[0067] The specific implementation of the present application is further described in detail below with reference to the accompanying drawings and examples.

[0068] This application designs a distributed control strategy for an AC NMG system, where multiple low-voltage MGs are connected in parallel to a medium-voltage grid. This strategy aims to improve the resilience and resiliency of the microgrid while further optimizing the network losses of the NMG system. Since this application focuses on an AC microgrid operating in an isolated island, local autonomy and precise power distribution between units must be considered, so droop control is used as the basic control. Compared to the control of a single MG system, the NMG system couples more electrical components and communication infrastructure, making the control approach more challenging.

[0069] This application proposes a multi-process-based distributed control method for microgrids, which adopts a hierarchical control concept to systematically solve the shortcomings of existing methods. First, by analogy with the droop control of distributed generation units (DG), droop control is proposed for each microgrid (MG), so that it can operate autonomously only through local measurements. Distributed consistency control is carried out in the secondary process of the microgrid (MG) to achieve frequency and voltage recovery as well as active power and reactive power sharing; in the tertiary process of the microgrid (MG), the alternating direction multiplier method (ADMM) algorithm is used to optimize the power flow distribution in the microgrid, so that the total network power loss is minimized. In addition, the event triggering mechanism is applied to the edge computing device to reduce the demand for communication resources while ensuring the stable operation of each process in the edge agent.

[0070] Example 1:

[0071] This embodiment proposes a multi-process based distributed control method for microgrids, such as Figure 1 、 Figure 2 、 Figure 3 、 Figure 4 Shown, including:

[0072] Step S1: establishing an edge agent for each microgrid in the networked microgrid, each edge agent including: a main process, a secondary process, and a tertiary process, the networked microgrid including multiple microgrids, each microgrid including multiple distributed power generation units;

[0073] In this embodiment, the networked microgrid (NMG) system is an interactive, complex cyber-physical system consisting of highly coupled electrical and communication systems. Generally, the NMG system can operate in an island or grid-connected mode. This embodiment studies an island NMG system, in which each MG is integrated by a set of dispatchable DGs connected to a common coupling point (PCC). The communication system in the NMG system includes a lower-layer communication network between DGs within each MG and an upper-layer communication network between MGs. In the communication network of the NMG system, reference frequency and voltage signals are generated by measuring the tie line information of the local MG and exchanging it with adjacent agents. The generated reference frequency and voltage signals are then sent to the distributed secondary control of a single MG so that all DGs can track this reference signal. An event trigger detector is set in the edge computing of the upper and lower communication layers. Only when the preset threshold of the trigger function is met, the state variables are uploaded and the control instructions are updated, thereby alleviating the communication pressure of the system and reducing the communication cost.

[0074] Step S1.1: NMG system's upper communication network: a diagram of the communication network between multiple MGs Indicates that is an edge set, is a set of nodes. For i=1,...,N, node i represents the i-th MG, and (i,j)∈ε means that the j-th MG can communicate with the i-th MG. The weighted adjacency matrix is ​​expressed as A = [a ij ]∈R N×N , where i≠j, a ij > 0. If (j,i)∈ε, a ij = 0. If all a ij =a ji , then the graph is an undirected graph. The degree matrix Δ is expressed as Δ=diag{d i}∈R N×N , Then, The Laplace matrix is ​​defined as L = Δ - A = [l ij ]∈R N×N Consider a command generator and an augmented multi-agent system with N agents. The relevant graph is defined as where={(0,i)|i=1,…,N}∪and Therefore, node 0 represents the command generator. For i=1,...,N, if the i-th MG can communicate with the command generator, then a i0 > 0, otherwise a i0 =0.

[0075] Step S1.2: Lower level communication network of NMG system: For a single MG, the communication network between multiple DGs within it is shown in Figure 1. Indicates that is an edge set, is a set of nodes. For k=1,...,M, node k represents the kth DG, It means that the kth DG can communicate with the hth DG. The weighted adjacency matrix of G is expressed as A = [a kh ]∈R M×M , where k≠h, a kh > 0. If a kh = 0. If all a kh =a hk , then the graph is an undirected graph.

[0076] Step S2: In the third-level process, the networked microgrid is optimized by using an alternating direction multiplier algorithm to minimize the total reactive power loss in the networked microgrid. The active power reference value and reactive power reference value of each microgrid are obtained and sent to the second-level process, including:

[0077] Step S2.1: Establish an optimal power flow problem with the control objective of minimizing the total reactive power loss in the networked microgrid;

[0078] In this embodiment, a three-level process is established based on the secondary process of the NMG system to optimize the operation of the NMG system with the goal of minimizing network power loss in the microgrid. Since the dynamics of frequency and voltage recovery are much faster, the values ​​of frequency and voltage can be regarded as steady-state values ​​through the three-level optimization. The three-level control solves the optimal power flow (OPF) problem and returns the actual active power and reactive power dispatch commands to the secondary control layer ( and ), the proposed control framework will follow the scheduling commands to make the system operate in the optimal state.

[0079] The main advantage of using ADMM is that it inherits the advantages of dual decomposition and augmented Lagrangian methods for constrained optimization. In the general decomposition structure, the objective and constraints are divided into K parts:

[0080]

[0081]

[0082] Among them, x i is a local variable, z is a global variable, is the local variable x i should be the ratio of variable z, is a local constraint, and K is the number of constraints.

[0083] The augmented Lagrangian for the problem of coupled constraints in general decomposition structures is given.

[0084]

[0085] Among them, λ i is the dual variable associated with the equality constraint, and ρ∈R is the Lagrange step size parameter.

[0086] The optimal power flow problem is established with the control objective of minimizing the total reactive power loss in the networked microgrid. The calculation formula is as follows:

[0087]

[0088] st:

[0089]

[0090]

[0091]

[0092]

[0093]

[0094] in, Expressed as is the node voltage vector; is The vector obtained by eliminating the elements irrelevant to the busbar i; H i Including MG i and the number of MGs connected to it; MG i The minimum value of active power, MG i The maximum value of active power, MG i The minimum value of reactive power, MG i The maximum value of reactive power; and is the load power at bus i; Minimum voltage threshold, is the maximum voltage threshold, the matrix and Expressed as:

[0095]

[0096] where Y = G + jB is the admittance matrix of the grid system, and G is the impedance matrix. and are matrices derived from z p and z q by writing the constraints in quadratic form.

[0097] Step S2.2: According to the number of buses, the optimal power flow problem is decomposed into multiple sub-problems;

[0098] In this embodiment, in order to solve the OPF problem in a distributed manner, the system in step S2.1 needs to be split into subsystems. The number of subsystems corresponds to the number of MGs. The total active power loss can be represented by decomposing the function into H parts:

[0099]

[0100] The sub-problem on bus i involves H i bus lines, and can be represented as:

[0101] If

[0102]

[0103]

[0104] If

[0105]

[0106]

[0107] where is a local variable. Therefore, the OPF problem can be expressed as a general consensus problem of general decomposition structure, and can be solved in a distributed manner by ADMM, where the coupling constraint is:

[0108]

[0109] where is a global variable, and represents the mapping of the set of related components to subsystem i.

[0110] Step S2.3: Solve each sub-problem to obtain the active power reference value and the reactive power reference value of each micro-grid, and send them to the secondary process.

[0111] In this embodiment, by solving each of the above sub-problems (subsystems), the active power reference value and the reactive power reference value of each micro-grid are finally obtained.

[0112] Step S3: In the secondary process, based on the local measurement signals, the measurement signals from the adjacent microgrids, and the active power reference and reactive power reference values ​​from the tertiary process, a multi-agent consensus algorithm is used to perform distributed consistency control to achieve frequency recovery of the microgrid, voltage recovery at the point of common coupling, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the main process.

[0113] Step S3.1: Set a trigger function for the communication between adjacent microgrids. When the value of the trigger function meets the preset threshold, obtain the voltage status of the adjacent microgrid. and frequency status Used to update the voltage and frequency status of the local microgrid and to propagate the voltage status of the local microgrid to neighboring microgrids. and frequency status

[0114] In this embodiment, during the communication between the upper and lower layers of the NMG system, information exchange occurs only at non-periodic triggering moments, which are determined by event triggering conditions. The event triggering error and event triggering conditions are generally expressed as follows:

[0115] e(t)=x(t k )-x(t)

[0116] ||e(t)||≥e T

[0117] Among them, t∈[t k ,t k+1 ), x(t) is the current system state variable, x(t k ) is the system state variable of the k-th trigger event. When the formula e(t)=x(t k )-x(t) calculated state measurement error e(t) is greater than the trigger threshold e T , then the current moment t is marked as the triggering moment.

[0118] Among them, the time series expression of the following event triggering in the edge computing of the upper and lower communication layers is:

[0119] t k =inf{t>t k-1,i |f i (t)≥0}

[0120] The trigger function is calculated as follows:

[0121]

[0122] Among them, f i (t) is the triggering function of the i-th microgrid at the t-th moment, σ i is the first setting coefficient, α i is the second setting coefficient, and 0<σ i <1,0<α i <1 / d i , d i is the in-degree of the i-th microgrid, ζ avg,i is the voltage mean or frequency mean of the i-th microgrid, e i (t) is the measurement error, which is calculated as follows:

[0123]

[0124] Among them, e i (t) is the measurement error of the i-th microgrid at the t-th moment, is the measured value of the i-th microgrid at the t-th moment, i.e., voltage or frequency, is the i-th microgrid at the t-th k The measured value of the state variable at time t k+1 For the tth k+1 time, t k For the tth k time.

[0125] When the trigger function f i When (t) is satisfied, the neighbor information is obtained to update the local control signal and the local state information is disseminated to the neighbors. Since the event-triggered secondary control only requires the neighbor information, the reduction of information transmission reduces the communication cost and improves the communication efficiency.

[0126] In this embodiment, the convergence of the proposed trigger function is demonstrated below. The closed-loop control system can be expressed as:

[0127]

[0128] Among them, ζ is the system state variable, representing frequency or voltage; is the updated variable; L is the gain matrix; e is the measurement error of the state variable.

[0129] If we can prove that It can prove the convergence and stability of the system near the equilibrium point.

[0130]

[0131] Expanding the above formula, we can get:

[0132]

[0133] Where N is the number of microgrids, e i is the measurement error of the i-th microgrid state variable, e j is the jth microgrid state variable measurement error, ζ avg,i is the mean voltage or frequency of the i-th microgrid.

[0134] Use the following inequality:

[0135]

[0136] Get the boundary values ​​of the inequality:

[0137]

[0138] Among them, d i is the in-degree of the i-th microgrid.

[0139] When the conditions shown by the boundary values ​​are met:

[0140]

[0141] Among them, σ i is the first setting coefficient, α i is the second setting coefficient, and 0<σ i <1,0<α i <1 / d i , d i is the in-degree of the i-th microgrid.

[0142] It can be further deduced that:

[0143]

[0144] This proves the convergence of the event-triggered control algorithm.

[0145] Step S3.2: In the secondary process, based on the local measurement signals, the measurement signals from the adjacent microgrids, and the active power reference and reactive power reference values ​​from the tertiary process, a multi-agent consensus algorithm is used to perform distributed consistency control to achieve frequency recovery of the microgrid, voltage recovery at the point of common coupling, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the primary process.

[0146] In this embodiment, similar to the droop control of distributed generation units (DGs) in a single microgrid (MG), droop control is proposed for each MG, allowing it to operate autonomously using only local measurements. In addition, a secondary process is designed based on the MG droop control, and a multi-agent consensus algorithm is used to achieve frequency recovery in the AC microgrid and power sharing between MGs to address the accuracy trade-off between power sharing and frequency regulation caused by droop inertia. The MG secondary process aims to achieve three goals: (1) frequency recovery; (2) bus voltage recovery at the point of common coupling (PCC); and (3) arbitrary power sharing between MGs.

[0147] In the secondary process, according to each microgrid's local measurement signal and the measurement signal from the adjacent microgrid, as well as the active power reference value and the reactive power reference value from the tertiary process, a consensus algorithm is used to perform distributed consistency control on the frequency and voltage of each microgrid to obtain a frequency control signal and a voltage control signal, including:

[0148] Step S3.2.1: Based on the active power reference value of each microgrid, a multi-agent consensus algorithm is used to perform distributed consistency control on the frequency of each microgrid to obtain a frequency control signal. The calculation formula is as follows:

[0149] Based on the local frequency and active power measurement signals of each microgrid and the frequency and active power measurement signals from the adjacent microgrids, as well as the active power reference value from the three-level process, a multi-agent consensus algorithm is used to perform distributed consistency control on the frequency of each microgrid to obtain the frequency control signal. The calculation formula is as follows:

[0150]

[0151] in, For microgrid MG i The frequency control signal, N is the total number of microgrids, g i MG i The pinning gain, MG i stands for Local Microgrid, MG j represents the adjacent microgrid, a ij For local microgrid MG i With adjacent microgrid MG j The weighted adjacency matrix of For local microgrid MG i The active power reference value of For adjacent microgrid MG j The active power reference value of To regulate MG i The droop coefficient of active power, To regulate MG jThe droop coefficient of active power, For local microgrid MG i The frequency measurement value, For adjacent microgrid MG j The frequency measurement value, For local microgrid MG i The active power measurement value of For adjacent microgrid MG j The active power measurement value of For the updated local microgrid MG i frequency control signal.

[0152] Step S3.2.2: Based on the local voltage signal and reactive power measurement signal of each microgrid and the voltage signal and reactive power measurement signal from the adjacent microgrid, as well as the reactive power reference value from the three-level process, a multi-agent consensus algorithm is used to perform distributed consistency control on the voltage of each microgrid to obtain the voltage control signal. The calculation formula is as follows:

[0153]

[0154] in, For microgrid MG i The voltage control signal, N is the total number of microgrids, MG i stands for Local Microgrid, MG j represents the adjacent microgrid, a ij For local microgrid MG i With adjacent microgrid MG j The weighted adjacency matrix of For local microgrid MG i The reactive power reference value, For adjacent microgrid MG j Reactive power reference value, V * is the voltage standard value, To regulate MG i The droop factor of reactive power, To regulate MG j The droop factor of reactive power, For adjacent microgrid MG j The voltage control signal, For local microgrid MG i The measured value of reactive power, For adjacent microgrid MG j The measured value of reactive power, V PCC is the voltage at the point of common coupling, For the updated local microgrid MG i voltage control signal.

[0155] Step S3.2.3: In the secondary process, a multi-agent consensus algorithm is used for distributed consistency control. The frequency and common coupling voltage of the microgrid will converge to the standard value, which is expressed as follows:

[0156]

[0157]

[0158]

[0159]

[0160] Among them, ω * is the standard value of frequency, V * is the voltage standard value, is the local microgrid MG at time t i Frequency, V PCC is the voltage at the point of common coupling, To regulate MG j The droop coefficient of active power, To regulate MG i The droop coefficient of active power, is the adjacent microgrid MG at time t j The measured value of active power, is the local microgrid MG at time t i The measured value of active power, For adjacent microgrid MG j The active power reference value of For local microgrid MG i The active power reference value of To regulate MG j The droop factor of reactive power, To regulate MG i The droop factor of reactive power, is the adjacent microgrid MG at time t j The measured value of reactive power, Q MGi (t) is the local microgrid MG at time t i The measured value of reactive power, For local microgrid MG i The reactive power reference value, For adjacent microgrid MG j The reactive power reference value.

[0161] Step S4: In the main process, droop control is adopted for each microgrid to achieve autonomous operation. According to the voltage control signal and frequency control signal from the secondary process, the voltage reference value and the frequency reference value are calculated and sent to the droop controller of the power generation unit.

[0162] In the main process, droop control is adopted for each microgrid to achieve autonomous operation. The calculation formula is as follows:

[0163]

[0164]

[0165] Among them, ω * is the standard value of frequency, V * is the voltage standard value, is the frequency control signal from the secondary process; is the voltage control signal from the secondary process; To regulate MG i The droop coefficient of active power, To regulate MG i The droop factor of reactive power, For microgrid MG i The calculated frequency value is used as the frequency reference value sent to the distributed generation unit and as the frequency measurement value in the secondary process. For microgrid MG i The calculated voltage value is used as the voltage reference value sent to the distributed generation unit and as the voltage measurement value in the secondary process. For microgrid MG i The measured value of active power, For microgrid MG i A measurement of reactive power.

[0166] Step S5: droop control is adopted for the distributed generation units in a single microgrid, and then quadratic linear control is performed to track the voltage reference value and frequency reference value fed into the main process, so as to achieve frequency recovery, voltage recovery of the distributed generation units in a single microgrid, and active power distribution and reactive power distribution among multiple generation units.

[0167] In this embodiment, for the control of a single microgrid (MG), droop control is applied to control the interface voltage and frequency of the DG converter of the power generation unit in the island operation mode. A quadratic linear control of the distributed generation unit is designed to track the frequency and voltage reference values ​​fed into the corresponding MG, thereby achieving frequency recovery and voltage recovery of the distributed generation units within a single MG, as well as active power distribution and reactive power distribution among multiple generation units.

[0168] Step S5.1: Apply droop control to the distributed generation unit in a single microgrid, that is, apply the droop control to control the interface voltage and frequency of the DG converter of the generation unit in the island operation mode, and the expression is as follows:

[0169]

[0170]

[0171] in, is the frequency calculation value of the power generation unit, is the calculated value of the voltage of the power generation unit, For the microgrid MG i The frequency reference value, For the microgrid MG i The voltage reference value, To adjust the power generation unit DG k The droop coefficient of active power, To adjust the power generation unit DG k The ranges of reactive power droop coefficient are:

[0172]

[0173]

[0174] in, To control the maximum frequency of the power generation unit, To control the minimum frequency of the generating unit, DG is the power generation unit k The maximum value of active power, DG is the power generation unit k The minimum value of active power, DG is the power generation unit k The maximum value of the voltage, DG is the power generation unit k The minimum value of the voltage, DG is the power generation unit k The maximum value of reactive power, DG is the power generation unit k Minimum value of reactive power.

[0175] Step S5.2: The distributed power generation unit is subjected to quadratic linear control, and the calculation formula is as follows:

[0176]

[0177]

[0178]

[0179]

[0180] in, For microgrid MG i Medium power generation unit DG k and power generation unit DG h Communication coefficient between the power generation unit DG k and power generation unit DG h If there is a link between otherwise For microgrid MG i Medium power generation unit DG k The restraining gain of the power generation unit DG k Can receive directly and but otherwise DG is the power generation unit k The calculated frequency value of To control the power generation unit DG h The calculated frequency value of Adjusting the power generation unit DG k The droop coefficient of active power, Adjusting the power generation unit DG h The droop coefficient of active power, DG is the power generation unit k The measured value of active power, DG is the power generation unit k The frequency control signal, For microgrid MG i The total number of power generation units in DG is the power generation unit h The measured value of active power, DG is the power generation unit k The calculated voltage value, DG is the power generation unit k The droop factor of reactive power, DG is the power generation unit k The measured value of reactive power, DG is the power generation unit k The voltage control signal, DG is the power generation unit h The voltage control signal, For the microgrid MG i The frequency reference value, For the microgrid MG i The voltage reference value, DG is the power generation unit k The calculated voltage value, DG is the power generation unit h The droop factor of reactive power, DG is the power generation unit h The measured value of reactive power, DG is the power generation unit k The measured value of reactive power, DG is the power generation unit k The droop factor of reactive power, DG is a distributed generation unit k an updated frequency control signal; DG is a distributed generation unit k Updated voltage control signal.

[0181] In this embodiment, secondary control is performed on the DG to track the frequency and voltage reference values ​​fed into the corresponding MG, thereby achieving accurate power distribution among the DGs within a single MG.

[0182] In order to describe the specific process in more detail, a specific example is given for the above-mentioned multi-process-based distributed control method for microgrids, and the implementation process is described in detail.

[0183] Step S100: Based on the distributed multi-layer control solution proposed for the MGs, an edge agent is established for each MG. Each agent contains three parallel and independent processes: a primary process, a secondary process, and a tertiary process. These three processes operate on different timescales. When the primary process completes a cycle, the secondary process stores the signal value of the primary process until the primary process receives a new signal. This cycle ultimately achieves the goal of distributed event-triggered control of the microgrid based on multiple processes using edge agents.

[0184] Step S100.1: In the main process, similar to the DG droop control, droop control is proposed for each MG so that it can operate autonomously with only local measurements.

[0185] Step S100.1.1: Obtain local measurement value of node i and

[0186] Step S100.1.2: Calculate the control signal according to the droop control and secondary control signal and

[0187] Step S100.1.3: and Send to MG i Local droop control;

[0188] Return to step S100.1.4: Return to step S100.1.1.

[0189] Step S100.2: The secondary control layer corresponds to the secondary process and uses event-triggered control. In the secondary process, a large amount of data (such as voltage and frequency) generated between adjacent MGs will be analyzed, processed, and stored in the edge computing device, with better real-time performance.

[0190] Step S100.2.1: Acquire On node MG i Local measurement value at ;

[0191] Step S100.2.2: If the event trigger function condition is met, proceed to the next step, i.e., step S100.2.3;

[0192] Step S100.2.3: Exchange collected local measurements

[0193] Step S100.2.4: Calculate control signals based on local measurements and three-level process signals and

[0194] Step S100.2.5: Send To the parent process;

[0195] Step S100.2.6: Return to step S100.2.1 and restart a new consensus algorithm cycle.

[0196] Step S100.3: The three-level control layer corresponds to the three-level process. The scheduling layer adopts time-triggered control with a sampling period of 1s. In the cloud microgrid control center, the ADMM algorithm is used to optimize the operation of the microgrid system, with the goal of minimizing the total network power loss in the microgrid, solving the optimal power flow problem, and returning the actual active power and reactive power scheduling commands to the secondary control layer ( and ). The proposed control framework will follow the scheduling commands to make the system operate in the optimal state.

[0197] Step S100.3.1: Start the loop and perform the initial iteration: I=1;

[0198] Step S100.3.2: Obtain the initial state, active power, and reactive power of the local measured load at node i:

[0199] Step S100.3.3: Initial value of global variables:

[0200] Step S100.3.4: Initial value of Lagrange multiplier: λ i (I)←λ0;

[0201] Step S100.3.5: When When , the loop starts to execute;

[0202] Step S100.3.6: Solve a local non-convex optimization problem to update the local variable v i (I+1). At the MGs bus, the local voltage is set to a specific reference value. The problem at the bus is:

[0203]

[0204]

[0205] Step S100.3.7:

[0206] Step S100.3.8: Distribute U to all neighbors;

[0207] Step S100.3.9: Collect U from all neighbors;

[0208] Step S100.3.10: Update the global variable by averaging all received U

[0209] Step S100.3.11: Update the Lagrange multipliers:

[0210] Step S100.3.12: Enter the next iteration: I=I+1;

[0211] Step S100.3.13: Satisfy Proceed to the next step;

[0212] Step S100.3.14: Calculate the set point power output for the corresponding DR:

[0213]

[0214]

[0215] Step S100.3.15: Sent to the secondary process, completing the current loop.

[0216] Step S100.3.16: Start a new ADMM cycle from step S100.3.1.

[0217] Simulation results:

[0218] This example uses four MG units as an example to simulate the frequency, voltage and active power. Figure 5 As shown, the traditional droop control method is used from 0 to 1s, and the control method proposed in this example is used from 1 to 10s. Figure 5 It can be seen that this example control method can effectively restore voltage and frequency and reasonably distribute power.

[0219] This embodiment proposes a multi-process based distributed control method for microgrids, including: for networked microgrid (NMG) systems, adopting the idea of ​​hierarchical control to coordinate control under multiple processes. An event trigger detector is set in the edge computing of the communication layer. Only when the preset threshold of the trigger function is met, the state variables are uploaded and the control instructions are updated, so as to achieve the purpose of reducing the number of communications and saving communication costs on the basis of ensuring the stable operation of the system. For the DG unit in a single MG, droop control is used for frequency and voltage control, and quadratic linear control is performed to track the voltage reference value and frequency reference value ( and ), realizing frequency recovery and voltage recovery of distributed generation units in a single microgrid, as well as active power distribution and reactive power distribution among multiple distributed generation units. For the main control layer of the NMG system, similar to the DG droop control, a droop control is proposed for each MG, so that it can operate autonomously only through local measurement, and the generated reference frequency and voltage ( and ) is sent to the droop controller of the local DG. The secondary control of the NMG system is to perform distributed consistency control through a multi-agent consensus algorithm to calculate the control signal based on the local measurement signal ( and ) and sends the control signal to the droop control in the master process, thereby restoring the frequency and common coupling point voltage in the NMG system to the reference value, and the active power and reactive power will follow the dispatch signal from the three-level control ( and The three-level control of the NMG system uses the ADMM algorithm to optimize the operation of the microgrid system, with the goal of minimizing the network power loss in the microgrid, solving the optimal power flow problem, and returning the actual active power and reactive power dispatch commands to the secondary control layer ( and The three control processes are integrated into a single MG edge agent, which follows scheduling commands and runs at different timescales. When the superior process completes a cycle, the subordinate process saves the superior process's signal value until the superior process receives a new sample. This cycle ultimately achieves the multiple goals of distributed event-triggered control of a microgrid based on multiple processes in the edge agent.

[0220] Example 2:

[0221] This embodiment proposes a microgrid distributed control system based on multi-process, such as Figure 6 As shown, it includes: an edge agent module, a three-level process module, a two-level process module, a main process module and a single microgrid control module, the edge agent module is connected to the three-level process module, the two-level process module and the main process module respectively, the three-level process module is connected to the two-level process module, the two-level process module is connected to the main process module, and the main process module is connected to the single microgrid control module;

[0222] An edge agent module establishes an edge agent for each microgrid in the networked microgrid, each edge agent including: a main process, a secondary process, and a tertiary process, the networked microgrid including multiple microgrids, each microgrid including multiple distributed power generation units;

[0223] The third-level process module is used to optimize the networked microgrid by adopting an alternating direction multiplier algorithm in the third-level process to minimize the total reactive power loss in the networked microgrid, obtain the active power reference value and the reactive power reference value of each microgrid, and send them to the second-level process;

[0224] The secondary process module is used to perform distributed consistency control in the secondary process using a multi-agent consensus algorithm based on local measurement signals, measurement signals from adjacent microgrids, and active power reference values ​​and reactive power reference values ​​from the tertiary process to achieve frequency recovery of the microgrid, voltage recovery at the common coupling point, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the main process.

[0225] The main process module is used to implement droop control on each microgrid in the main process to achieve autonomous operation. It calculates the voltage reference value and frequency reference value based on the voltage control signal and frequency control signal from the secondary process, and sends the voltage reference value and frequency reference value to the droop controller of the power generation unit.

[0226] A single microgrid control module is used to adopt droop control on the distributed generation units in a single microgrid, and then perform quadratic linear control to track the voltage reference value and frequency reference value fed into the main process, so as to achieve frequency recovery, voltage recovery of the distributed generation units in the single microgrid, and active power distribution and reactive power distribution among multiple generation units.

[0227] The various embodiments in this application are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0228] The scope of the present application is not limited to the above-described embodiments, and it will be apparent to those skilled in the art that various changes and modifications can be made to the present disclosure without departing from the scope and spirit of the present disclosure. The present disclosure is intended to include such changes and modifications within the scope and spirit of the claims of the present disclosure and their equivalents.

Claims

1. A multi-process based distributed control method for microgrids, characterized in that: include: Establishing an edge agent for each microgrid in a networked microgrid, each edge agent comprising: a main process, a secondary process, and a tertiary process, wherein the networked microgrid comprises a plurality of microgrids, each microgrid comprising a plurality of distributed generation units; In the third-level process, the alternating direction multiplier algorithm is used to optimize the networked microgrid with the goal of minimizing the total reactive power loss in the networked microgrid. The active power reference value and reactive power reference value of each microgrid are obtained and sent to the second-level process. In the secondary process, a multi-agent consensus algorithm is used for distributed consistency control based on local measurement signals, measurement signals from adjacent microgrids, and active and reactive power reference values ​​from the tertiary process. This allows for frequency recovery of the microgrid, voltage recovery at the point of common coupling, and arbitrary power sharing between microgrids. The calculated voltage and frequency control signals are then sent to the droop controller in the primary process. In the main process, droop control is adopted for each microgrid to achieve autonomous operation. According to the voltage control signal and frequency control signal from the secondary process, the voltage reference value and frequency reference value are calculated and sent to the droop controller of the power generation unit. The distributed generation units in a single microgrid are controlled by droop control and then by quadratic linear control to track the voltage reference value and frequency reference value fed into the main process, so as to achieve frequency recovery and voltage recovery of the distributed generation units in the single microgrid, as well as active power distribution and reactive power distribution among multiple generation units. The multi-process-based microgrid distributed control method further includes: setting a trigger function for communication with adjacent microgrids, and when the value of the trigger function meets a preset threshold, obtaining the voltage state and frequency state of the adjacent microgrid for updating the voltage state and frequency state of the local microgrid, and simultaneously propagating the voltage state and frequency state of the local microgrid to the adjacent microgrid. The trigger function is calculated as follows: Among them, f i (t) is the trigger function of the i-th microgrid at the t-th moment, σ i is the first setting coefficient, α i is the second setting coefficient, and 0<σ i <1,0<α i <1 / d i , d i is the in-degree of the i-th microgrid, ζ avg,i is the mean value of the voltage or frequency of the i-th microgrid, e i (t) is the measurement error, which is calculated as follows: Among them, e i (t) is the measurement error of the i-th microgrid at the t-th moment, is the measured value of the i-th microgrid at the t-th moment, i.e., voltage or frequency, is the tth microgrid k The measured value at time t k+1 For the tth k+1 time, t k For the tth k time.

2. The multi-process based microgrid distributed control method according to claim 1, characterized in that: The alternating direction multiplier algorithm is used to optimize the networked microgrid with minimizing the total reactive power loss in the networked microgrid as the control goal, and the active power reference value and reactive power reference value of each microgrid are obtained, including: The optimal power flow problem is established with the control objective of minimizing the total reactive power loss in the networked microgrid; Decompose the optimal power flow problem into multiple sub-problems according to the number of buses; Solve each sub-problem to obtain the active power reference value and reactive power reference value of each microgrid and send them to the secondary process.

3. The multi-process based microgrid distributed control method according to claim 1, characterized in that: In the secondary process, a multi-agent consensus algorithm is used to perform distributed consistency control based on local measurement signals, measurement signals from adjacent microgrids, and active power reference values ​​and reactive power reference values ​​from the tertiary process to achieve frequency recovery of the microgrid, voltage recovery at the point of common coupling, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the main process, including: Based on the local frequency measurement signal and active power measurement signal of each microgrid, the frequency measurement signal and active power measurement signal from the adjacent microgrid, and the active power reference value from the three-level process, a multi-agent consensus algorithm is used to perform distributed consistency control on the frequency of each microgrid to obtain the frequency control signal. The calculation formula is as follows: in, For microgrid MG i The frequency control signal, N is the total number of microgrids, g i MG i The pinning gain, MG i stands for Local Microgrid, MG j represents the adjacent microgrid, a ij For local microgrid MG i With adjacent microgrid MG j The weighted adjacency matrix of For local microgrid MG i The active power reference value of For adjacent microgrid MG j The active power reference value of To adjust MG i The droop coefficient of active power, To adjust MG j The droop coefficient of active power, For local microgrid MG i The frequency measurement value, For adjacent microgrid MG j The frequency measurement value, For local microgrid MG i The active power measurement value of For adjacent microgrid MG j The measured active power value of For the updated local microgrid MG i The frequency control signal, ω * is the standard value of frequency; Based on the local voltage measurement signal and reactive power measurement signal of each microgrid and the voltage measurement signal and reactive power measurement signal from the adjacent microgrid and the reactive power reference value from the three-level process, a multi-agent consensus algorithm is used to perform distributed consistency control on the voltage of each microgrid to obtain the voltage control signal. The calculation formula is as follows: in, For microgrid MG i The voltage control signal, N is the total number of microgrids, MG i stands for Local Microgrid, MG j represents the adjacent microgrid, a ij For local microgrid MG i With adjacent microgrid MG j The weighted adjacency matrix of For local microgrid MG i The reactive power reference value, For adjacent microgrid MG j Reactive power reference value, V * is the voltage standard value, To adjust MG i The droop factor of reactive power, To adjust MG j The droop factor of reactive power, For adjacent microgrid MG j The voltage control signal, For local microgrid MG i The measured value of reactive power, For adjacent microgrid MG j The measured value of reactive power, V PCC is the voltage at the point of common coupling, For the updated local microgrid MG i voltage control signal.

4. The multi-process based microgrid distributed control method according to claim 3, characterized in that: In the secondary process, a multi-agent consensus algorithm is used for distributed consistency control, and the frequency and common coupling voltage of the microgrid will converge to the standard value, which is expressed as follows: Among them, ω * is the standard value of frequency, V * is the voltage standard value, is the local microgrid MG at time t i Frequency, V PCC is the voltage at the point of common coupling, To adjust MG j The droop coefficient of active power, To adjust MG i The droop coefficient of active power, is the adjacent microgrid MG at time t j The measured value of active power, is the local microgrid MG at time t i The measured value of active power, For adjacent microgrid MG j The active power reference value of For local microgrid MG i The active power reference value of To adjust MG j The droop factor of reactive power, To adjust MG i The droop factor of reactive power, is the adjacent microgrid MG at time t j The measured value of reactive power, is the local microgrid MG at time t i The measured value of reactive power, For local microgrid MG i The reactive power reference value, For adjacent microgrid MG j The reactive power reference value.

5. The multi-process based microgrid distributed control method according to claim 3, characterized in that: In the main process, droop control is adopted for each microgrid to achieve autonomous operation. The calculation formula is as follows: Among them, ω * is the standard value of frequency, V * is the voltage standard value, is the frequency control signal from the secondary process; is the voltage control signal from the secondary process; To adjust MG i The droop coefficient of active power, To adjust MG i The droop factor of reactive power, For microgrid MG i The calculated frequency value is used as the frequency reference value sent to the distributed generation unit and as the frequency measurement value in the secondary process. For microgrid MG i The calculated voltage value is used as the voltage reference value sent to the distributed generation unit and as the voltage measurement value in the secondary process. For microgrid MG i The measured value of active power, For microgrid MG i A measurement of reactive power.

6. The multi-process based microgrid distributed control method according to claim 1, characterized in that: The distributed generation unit in a single microgrid adopts droop control, and the calculation formula is as follows: in, is the frequency calculation value of the power generation unit, is the calculated value of the voltage of the power generation unit, For the microgrid MG i The frequency reference value, For the microgrid MG i The voltage reference value, To adjust the power generation unit DG k The droop coefficient of active power, To adjust the power generation unit DG k The ranges of reactive power droop coefficient are: in, To control the maximum frequency of the power generation unit, To control the minimum frequency of the generating unit, DG is the power generation unit k The maximum value of active power, DG is the power generation unit k The minimum value of active power, DG is the power generation unit k The maximum value of the voltage, DG is the power generation unit k The minimum value of the voltage, DG is the power generation unit k The maximum value of reactive power, DG is the power generation unit k Minimum value of reactive power.

7. The multi-process based microgrid distributed control method according to claim 6, characterized in that: The quadratic linear control is performed on the distributed power generation unit, and the calculation formula is as follows: in, For microgrid MG i Medium power generation unit DG k and power generation unit DG h Communication coefficient between the power generation unit DG k and power generation unit DG h If there is a link between otherwise For microgrid MG i Medium power generation unit DG k The restraining gain of the power generation unit DG k Can receive directly and but otherwise DG is the power generation unit k The calculated frequency value of To control the power generation unit DG h The calculated frequency value of Adjusting the power generation unit DG k The droop coefficient of active power, Adjusting the power generation unit DG h The droop coefficient of active power, DG is the power generation unit k The measured value of active power, DG is the power generation unit k The frequency control signal, For microgrid MG i The total number of power generation units in DG is the power generation unit h The measured value of active power, DG is the power generation unit k The calculated voltage value, DG is the power generation unit k The droop factor of reactive power, DG is the power generation unit k The measured value of reactive power, DG is the power generation unit k The voltage control signal, DG is the power generation unit h The voltage control signal, For the microgrid MG i The frequency reference value, For the microgrid MG i The voltage reference value, DG is the power generation unit k The calculated voltage value, DG is the power generation unit h The droop factor of reactive power, DG is the power generation unit h The measured value of reactive power, DG is the power generation unit k The measured value of reactive power, DG is the power generation unit k The droop factor of reactive power, DG is a distributed generation unit k an updated frequency control signal; DG is a distributed generation unit k Updated voltage control signal.

8. A multi-process based microgrid distributed control system, implemented by the multi-process based microgrid distributed control method according to any one of claims 1 to 7, characterized in that: include: An edge agent module, a three-level process module, a two-level process module, a main process module, and a single microgrid control module, wherein the edge agent module is connected to the three-level process module, the two-level process module, and the main process module respectively, the three-level process module is connected to the two-level process module, the two-level process module is connected to the main process module, and the main process module is connected to the single microgrid control module; An edge agent module is used to establish an edge agent for each microgrid in the networked microgrid, each edge agent including: a main process, a secondary process and a tertiary process, the networked microgrid including multiple microgrids, each microgrid including multiple distributed power generation units; The third-level process module is used to optimize the networked microgrid by adopting an alternating direction multiplier algorithm in the third-level process to minimize the total reactive power loss in the networked microgrid, obtain the active power reference value and the reactive power reference value of each microgrid, and send them to the second-level process; The secondary process module is used to perform distributed consistency control in the secondary process using a multi-agent consensus algorithm based on local measurement signals, measurement signals from adjacent microgrids, and active power reference values ​​and reactive power reference values ​​from the tertiary process to achieve frequency recovery of the microgrid, voltage recovery at the common coupling point, and arbitrary power sharing between microgrids. The calculated voltage control signal and frequency control signal are sent to the droop controller in the main process. The main process module is used to implement droop control for each microgrid in the main process to achieve autonomous operation. The voltage reference value and frequency reference value are calculated based on the voltage control signal and frequency control signal from the secondary process, and the voltage reference value and frequency reference value are sent to the droop controller of the power generation unit. A single microgrid control module is used to adopt droop control on the distributed generation units in a single microgrid, and then perform quadratic linear control to track the voltage reference value and frequency reference value fed into the main process, so as to achieve frequency recovery, voltage recovery of the distributed generation units in the single microgrid, and active power distribution and reactive power distribution among multiple generation units.

Images