A direct current micro-grid voltage control method and computer readable medium

Abstract

The application provides a DC micro-grid voltage control method and a computer readable medium. The application inputs load and line data of a DC micro-grid, and constructs a DC micro-grid model; secondly, considering the uncertainty of the system, a state equation of the DC micro-grid model is constructed, a dynamic matrix representing the state of the grid is calculated, a voltage control optimization model of the DC micro-grid model is constructed by using eigenvalue analysis, and corresponding constraint conditions are designed; finally, an improved particle swarm algorithm is used to solve the optimization model, a global optimal solution is output, optimal control parameters of the voltage controller under the condition of system condition change are determined, and real-time control of the voltage of the DC micro-grid is realized. The application has the advantages that the eigenvalue analysis is used to define an objective function of the optimization model, local optimization and global optimization are realized in parallel through different search modes of search particles, and therefore the parameters of the voltage controller can be updated in real time under the condition of system condition change.

Description

Technical Field

[0001] This invention relates to the field of DC microgrid voltage control, and more specifically to a DC microgrid voltage control method and a computer-readable medium. Technical Background

[0002] With the increasing integration of renewable distributed energy sources such as solar and wind power into the power grid, the stability problems caused by their intermittency and volatility are becoming increasingly serious. Furthermore, as network complexity increases, traditional control methods such as energy storage systems are becoming increasingly difficult to address these issues. DC microgrids are gaining attention due to their high power supply reliability, high energy utilization, and the absence of reactive power and frequency issues associated with AC grids. However, DC microgrids are susceptible to voltage fluctuations caused by uncertainties such as converter interference, distributed power source switching, and load changes. Therefore, a voltage control method is needed to eliminate these fluctuations.

[0003] Traditional feedback control struggles to achieve stable voltage control quickly under drastic changes in system conditions. This invention proposes a DC microgrid voltage control method that considers system uncertainties. It transforms the controller parameter optimization problem into a dynamic matrix eigenvalue optimization problem and utilizes the high-speed optimization capability of the particle swarm optimization algorithm to update control parameters in real time, effectively ensuring the stability of voltage control. Summary of the Invention:

[0004] To address the aforementioned technical problems, this invention proposes a DC microgrid voltage control method and a computer-readable medium.

[0005] The technical solution of this invention is a DC microgrid voltage control method, and the specific steps are as follows:

[0006] Step 1: Input the load and line data of the DC microgrid to build a DC microgrid model;

[0007] Step 2: Construct the state equations of the DC microgrid model, calculate the dynamic matrix of the DC microgrid model, calculate the eigenvalues ​​of the dynamic matrix of the DC microgrid model, and use the eigenvalue analysis of the dynamic matrix of the DC microgrid model to construct the voltage control optimization model of the DC microgrid model.

[0008] Step 3: Construct the node voltage constraints, branch current constraints, and voltage control parameter constraints of the distributed units of the DC microgrid model respectively.

[0009] Step 4: Solve the DC microgrid voltage control optimization model using the improved particle swarm optimization algorithm. Set the initial parameters of the improved particle swarm optimization algorithm, construct the fitness function by combining the eigenvalue parameters of the dynamic matrix of the DC microgrid model, and select the proportional gain and differential gain of the voltage controller of the distributed unit of the DC microgrid model as the decision variables of the optimization algorithm.

[0010] Step 5: Using the position vector elements of each particle at the initial iteration as the initial proportional gain and initial differential gain of the voltage controller of each distributed unit in the DC microgrid model, perform power flow calculation on the DC microgrid model to obtain the voltage of each node and the current of each branch in the DC microgrid model, and update the dynamic matrix and its eigenvalue parameters of the DC microgrid model accordingly. Update the position vector and velocity vector of the regular particles, update the position vector of the random particles, calculate the fitness function value corresponding to the position vector of each particle according to the eigenvalue parameters of the dynamic matrix obtained from the power flow calculation, and update the local optimal solution and global optimal solution of each particle within the constraints.

[0011] Step 6: Repeat step 5 until the maximum number of iterations is reached to obtain the global optimal solution. Based on the global optimal solution, obtain the optimal proportional gain and optimal differential gain of the voltage controller of each distributed unit in the DC microgrid model to realize real-time control of the DC microgrid voltage.

[0012] Preferably, the DC microgrid model constructed in step 1 includes N nodes, and M nodes are selected from the N nodes and incorporated into the distributed unit;

[0013] Preferably, the state equation of the DC microgrid model described in step 2 is:

[0014]

[0015] m = 1, 2, ..., N

[0016] n = 1, 2, ..., N

[0017] m≠n

[0018] in, Let be the derivative of the output current of the distributed unit connected to the m-th node in the DC microgrid model at time t. Let be the derivative of the node voltage of the m-th node in the DC microgrid model at time t. Let ΔI be the derivative of the branch current between the m-th node and the n-th node in the DC microgrid model at time t. DG,m (t) represents the difference between the output current of the distributed unit connected to the m-th node in the DC microgrid model at time t and the previous sampling time, ΔU m(t) represents the difference between the node voltage of the m-th node in the DC microgrid model at time t and the previous sampling time, ΔI. m,n (t) represents the difference between the branch current between the m-th node and the n-th node in the DC microgrid model at time t and the previous sampling time, where m is the number of the m-th node in the DC microgrid model, n is the number of the n-th node in the DC microgrid model, N is the total number of nodes in the DC microgrid model, t represents the t-th time, Δt is the sampling time, and L DG,m Let C be the equivalent series inductance of the distributed unit connected to the m-th node in the DC microgrid model. m Let L be the equivalent capacitance connected in parallel at the m-th node in the DC microgrid model. m,n Let R be the branch inductance between the m-th node and the n-th node in the DC microgrid model. m,n Let E be the branch resistance between the m-th node and the n-th node in the DC microgrid model. m D is the equivalent DC power supply for the distributed unit connected to the m-th node in the DC microgrid model. m (t) represents the duty cycle of the buck chopper of the distributed unit connected to the m-th node in the DC microgrid model at time t. m (t) represents the node voltage of the m-th node in the DC microgrid model at time t, U n (t) represents the node voltage of the nth node in the DC microgrid model at time t. DG,m (t) represents the output current of the distributed unit connected to the m-th node in the DC microgrid model at time t. LD,m (t) represents the load current at the m-th node at time t, I m,n (t) represents the branch current between the m-th node and the n-th node in the DC microgrid model at time t;

[0019] Take I DG,m U m I m,n As a state variable, with I LD,m The state equations of the DC microgrid model are linearized using algebraic variables to obtain the dynamic matrix of the DC microgrid model.

[0020] The eigenvalues ​​of the dynamic matrix of the DC microgrid model described in step 2 are:

[0021] λ r =δ r +jω r

[0022] r = 1, 2, ..., T

[0023] Where, λ rLet δ be the r-th eigenvalue of the dynamic matrix of the DC microgrid model, T be the number of eigenvalues ​​of the dynamic matrix of the DC microgrid model, and δ be the eigenvalue of the dynamic matrix of the DC microgrid model. r ω is the real part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r This represents the imaginary part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model.

[0024] The voltage control optimization model for the DC microgrid model described in step 2 is defined as follows:

[0025]

[0026] Where T is the number of eigenvalues ​​in the dynamic matrix of the DC microgrid model, δ r ω is the real part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r Let a be the imaginary part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r b is the damping weighting coefficient for the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r ω is the angular frequency weighting coefficient of the r-th eigenvalue of the dynamic matrix of the DC microgrid model;

[0027] Preferably, the node voltage constraint of the DC microgrid model in step 3 is:

[0028] U min ≤U m ≤U max

[0029] Among them, U m Let U be the node voltage of the m-th node in the DC microgrid model. min U represents the minimum node voltage in the DC microgrid model. max This represents the maximum value of the node voltage in the DC microgrid model;

[0030] The branch current constraints of the DC microgrid model described in step 3 are as follows:

[0031] |I m,n |≤I max

[0032] Among them, I m,n Let I be the branch current between the m-th node and the n-th node in the DC microgrid model. max The current flowing through the branch when the line reaches its maximum heat tolerance;

[0033] The voltage control parameter constraints for the distributed unit of the DC microgrid model described in step 3 are as follows:

[0034] k p,min ≤k p ≤k p,max

[0035] k d,min ≤k d ≤k d,max

[0036] Where, k p k represents the proportional gain of the distributed unit voltage controller in the DC microgrid model. p,min k is the minimum proportional gain of the distributed unit voltage controller. p,max k represents the maximum proportional gain of the distributed unit voltage controller. d Let k be the differential gain of the distributed unit voltage controller in the DC microgrid model. d,min k is the minimum value of the differential gain of the distributed unit voltage controller. d,max This represents the maximum value of the differential gain of the distributed unit voltage controller;

[0037] Preferably, the initial parameters for setting the improved particle swarm optimization algorithm in step 4 are as follows:

[0038] Initialize the particle search space to have a dimension of 2M, a number of regular particles of D1, a number of random particles of D2, and a maximum number of iterations of k. max The inertial weight coefficient of the improved particle swarm algorithm is w, the local weight coefficient of the improved particle swarm algorithm is c1, and the global weight coefficient of the improved particle swarm algorithm is c2.

[0039] The initial position vector for each particle during the initialization iteration is defined as follows:

[0040]

[0041] j = 1, 2, ..., 2M-1, 2M

[0042] i = 1, 2, ..., D1, D1+1, ..., D1+D2

[0043] in, Let i be the position vector of particle i during the initial iteration, which is the position vector when the iteration number k = 0, and corresponds to the initial proportional gain and initial differential gain of the voltage controller of each distributed unit in the DC microgrid model. Let be the j-th element of the position vector of particle i during the initialization iteration. If j ≤ M, then it corresponds to the initial proportional gain of the voltage controller of the j-th distributed unit in the DC microgrid model. If j > M, then it corresponds to the initial differential gain of the voltage controller of the jM-th distributed unit in the DC microgrid model. 2M is the dimension of the particle search space, which corresponds to the total number of control parameters of the voltage controller of the distributed unit in the DC microgrid. D1 is the number of regular particles, and D2 is the number of random particles.

[0044] The elements of the position vector for each particle during the initialization iteration are defined as follows:

[0045]

[0046] j = 1, 2, ..., M

[0047]

[0048] j = M+1, M+2, ..., 2M

[0049] in, Let be the j-th element of the position vector of particle i during the initial iteration, 2M be the dimension of the particle search space, rand(0,1) be a random variable in the range (0,1), and k be the position vector of particle i. p,min k is the minimum proportional gain of the distributed unit voltage controller. p,max k represents the maximum proportional gain of the distributed unit voltage controller. d,min k is the minimum value of the differential gain of the distributed unit voltage controller. d,max This represents the maximum value of the differential gain of the distributed unit voltage controller;

[0050] The initial velocity vector for a regular particle during initialization iteration is defined as follows:

[0051]

[0052] j = 1, 2, ..., 2M-1, 2M

[0053] Among them, V i 0 Let i be the velocity vector of particle i during the initialization iteration. Let be the j-th element of the velocity vector of particle i during the initial iteration, and 2M be the dimension of the particle search space;

[0054] The elements of the velocity vector during the initialization iteration of a regular particle are defined as follows:

[0055]

[0056] j = 1, 2, ..., 2M-1, 2M

[0057] in, Let V be the j-th element of the velocity vector of particle i during the initial iteration, 2M be the dimension of the particle search space, rand(0,1) be a random variable in the range (0,1), and V be the velocity vector of particle i during the initial iteration. min To improve the minimum particle velocity in the particle swarm optimization algorithm, V max To improve the maximum particle velocity in the particle swarm optimization algorithm;

[0058] The fitness function mentioned in step 4 is:

[0059]

[0060] Where T is the number of eigenvalues ​​in the dynamic matrix of the DC microgrid model, δ r ω is the real part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r Let a be the imaginary part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r b is the damping weighting coefficient for the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r ω is the angular frequency weighting coefficient of the r-th eigenvalue of the dynamic matrix of the DC microgrid model;

[0061] As a preferred embodiment, the updating of the position and velocity vectors of the conventional particles in step 5 is specifically as follows:

[0062]

[0063] i = 1, 2, ..., D1

[0064] in, Let i be the position vector of particle i after the k-th iteration. V is the position vector of particle i after the (k+1)th iteration. i k Let V be the velocity vector of particle i after the k-th iteration. i k+1 Let be the velocity vector of particle i after the (k+1)th iteration, rand(0,1) be a random variable in the range (0,1), w be the inertia weight coefficient of the improved particle swarm optimization algorithm, c1 be the local weight coefficient of the improved particle swarm optimization algorithm, and c2 be the global weight coefficient of the improved particle swarm optimization algorithm. Let be the local optimal position vector of particle i after the k-th iteration. Let D1 be the globally optimal position vector after the k-th iteration, and D1 be the number of regular particles.

[0065] Step 5, updating the position vector of the random particle, is as follows:

[0066]

[0067] i = D1+1, D1+2, ..., D1+D2

[0068] in, Let i be the position vector of particle i after the (k+1)th iteration. All are random variables in the range (0,1) at the k-th iteration, k p,maxk represents the maximum proportional gain of the distributed unit voltage controller. d,max D1 represents the maximum value of the differential gain of the distributed unit voltage controller, D2 represents the number of regular particles, and D1 represents the number of random particles.

[0069] Compare the fitness function values ​​f corresponding to the updated position vectors of each particle. i k+1 The fitness function value f of the local optimal solution vector of each particle. i k* Update the local optimal solution vector for each particle:

[0070]

[0071] i = 1, 2, ..., D1 + D2

[0072] in, Let be the local optimal position vector of particle i after the (k+1)th iteration. Let be the position vector of particle i after the (k+1)th iteration. Let f be the local optimal position vector of particle i after the kth iteration. i k+1 Let f be the fitness function value of the position vector of particle i after k+1 iterations. i k* Let be the fitness function value of the local optimal solution vector of particle i after the k-th iteration;

[0073] The global optimal solution is selected and updated from the local optimal solutions, and its value is as follows:

[0074]

[0075] j = 1, 2, ..., 2M

[0076] g = m i inf i k+1* i = 1, 2, ..., D1 + D2

[0077] in, This is the vector of the globally optimal solution after the (k+1)th iteration. Let f be the j-th element of the global optimal solution vector after the (k+1)-th iteration, g be the number of the particle that minimizes the fitness function value after the (k+1)-th iteration, and f be the element of the vector. i k+1* , where is the fitness function value of the local optimal solution vector after the (k+1)th iteration of the i-th particle, 2M is the dimension of the particle search space, which is the total number of control parameters of the voltage controller of the distributed unit of the DC microgrid, D1 is the number of regular particles, and D2 is the number of random particles.

[0078] Preferably, the globally optimal solution described in step 6 is:

[0079] X g =[X g,1 ,X g,2 ,...X g,j ,...X g,2M ]

[0080] j = 1, 2, ..., 2M

[0081] Among them, X g To utilize the globally optimal solution obtained by the improved particle swarm optimization algorithm, which corresponds to the optimal proportional gain and optimal differential gain of the voltage controller of each distributed unit in the DC microgrid model, X g,j is the j-th element of the global optimal solution vector, and 2M is the dimension of the particle search space, which corresponds to the total number of control parameters of the voltage controller of the distributed unit of the DC microgrid;

[0082] Step 6 describes obtaining the global optimal solution using the optimization algorithm, which yields the optimal proportional gain and optimal derivative gain of each distributed unit voltage controller, as follows:

[0083]

[0084] l = 1, 2, ..., M

[0085] Where, k p,l Let k be the proportional gain of the voltage controller of the l-th distributed unit in the DC microgrid model. d,l Let X be the differential gain of the voltage controller of the l-th distributed unit in the DC microgrid model. g,l is the l-th element of the global optimal solution vector, and M is the number of distributed units in the DC microgrid model;

[0086] Step 6 describes the implementation of real-time voltage stability control for the DC microgrid, as follows:

[0087] The optimal proportional gain and optimal derivative gain of the voltage controller of each distributed unit are allocated to the generator sets of each distributed unit to achieve real-time voltage control of the DC microgrid.

[0088] The present invention also provides a computer-readable medium storing a computer program executed by an electronic device, which, when run on the electronic device, causes the steps of the DC microgrid voltage control method to be executed.

[0089] The advantage of this invention is that it transforms the controller parameter optimization problem into a dynamic matrix eigenvalue optimization problem, and uses the high-speed optimization capability of the improved particle swarm optimization algorithm to update the control parameters in real time, effectively ensuring the stability of voltage control. Attached Figure Description

[0090] Figure 1 : Flowchart of the method according to an embodiment of the present invention. Detailed implementation method:

[0091] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0092] In specific implementation, the method proposed in the technical solution of this invention can be automatically executed by those skilled in the art using computer software technology. System devices for implementing the method, such as computer-readable storage media storing the corresponding computer program of the technical solution of this invention and computer equipment including the computer program running the corresponding computer program, should also be within the protection scope of this invention.

[0093] The following is combined Figure 1 The technical solution of the method in the embodiments of the present invention is a DC microgrid voltage control method, as detailed below:

[0094] Step 1: Input the load and line data of the DC microgrid to build a DC microgrid model;

[0095] Step 1 describes the construction of a DC microgrid model, which includes N=6 nodes. M=3 nodes are selected from these N nodes and incorporated into the distributed unit.

[0096] Step 2: Construct the state equations of the DC microgrid model, calculate the dynamic matrix of the DC microgrid model, calculate the eigenvalues ​​of the dynamic matrix of the DC microgrid model, and use the eigenvalue analysis of the dynamic matrix of the DC microgrid model to construct the voltage control optimization model of the DC microgrid model.

[0097] The state equation of the DC microgrid model described in step 2 is:

[0098]

[0099] m = 1, 2, ..., N

[0100] n = 1, 2, ..., N

[0101] m≠n

[0102] in, Let be the derivative of the output current of the distributed unit connected to the m-th node in the DC microgrid model at time t. Let be the derivative of the node voltage of the m-th node in the DC microgrid model at time t. Let ΔI be the derivative of the branch current between the m-th node and the n-th node in the DC microgrid model at time t. DG,m (t) represents the difference between the output current of the distributed unit connected to the m-th node in the DC microgrid model at time t and the previous sampling time, ΔU m (t) represents the difference between the node voltage of the m-th node in the DC microgrid model at time t and the previous sampling time, ΔI. m,n (t) represents the difference between the branch current between the m-th node and the n-th node in the DC microgrid model at time t and the previous sampling time, where m is the number of the m-th node in the DC microgrid model, n is the number of the n-th node in the DC microgrid model, N is the total number of nodes in the DC microgrid model, t represents the t-th time, Δt is the sampling time, and L DG,m Let C be the equivalent series inductance of the distributed unit connected to the m-th node in the DC microgrid model. m Let L be the equivalent capacitance connected in parallel at the m-th node in the DC microgrid model. m,n Let R be the branch inductance between the m-th node and the n-th node in the DC microgrid model. m,n Let E be the branch resistance between the m-th node and the n-th node in the DC microgrid model. m D is the equivalent DC power supply for the distributed unit connected to the m-th node in the DC microgrid model. m (t) represents the duty cycle of the buck chopper of the distributed unit connected to the m-th node in the DC microgrid model at time t. m (t) represents the node voltage of the m-th node in the DC microgrid model at time t, U n (t) represents the node voltage of the nth node in the DC microgrid model at time t. DG,m (t) represents the output current of the distributed unit connected to the m-th node in the DC microgrid model at time t. LD,m (t) represents the load current at the m-th node at time t, I m,n (t) represents the branch current between the m-th node and the n-th node in the DC microgrid model at time t;

[0103] Take I DG,m U m I m,n As a state variable, with I LD,m The state equations of the DC microgrid model are linearized using algebraic variables to obtain the dynamic matrix of the DC microgrid model.

[0104] The eigenvalues ​​of the dynamic matrix of the DC microgrid model described in step 2 are:

[0105] λ r =δ r +jω r

[0106] r = 1, 2, ..., T

[0107] Where, λ r Let δ be the r-th eigenvalue of the dynamic matrix of the DC microgrid model, T = 14, and δ be the number of eigenvalues ​​in the dynamic matrix of the DC microgrid model. r ω is the real part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r This represents the imaginary part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model.

[0108] The voltage control optimization model for the DC microgrid model described in step 2 is defined as follows:

[0109]

[0110] Where T = 14 is the number of eigenvalues ​​in the dynamic matrix of the DC microgrid model, δ r ω is the real part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r Let a be the imaginary part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r b is the damping weighting coefficient for the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r The angular frequency weighting coefficient is the r-th eigenvalue of the dynamic matrix of the DC microgrid model. The damping weighting coefficient a and the angular frequency weighting coefficient b of the dynamic matrix of the DC microgrid model need to be given in combination with the size of the real and imaginary parts of the eigenvalue. The larger the real part of the eigenvalue, the larger the corresponding damping weighting coefficient a; the larger the imaginary part of the eigenvalue, the larger the corresponding angular frequency weighting coefficient b.

[0111] Step 3: Construct the node voltage constraints, branch current constraints, and voltage control parameter constraints of the distributed units of the DC microgrid model respectively.

[0112] The node voltage constraints of the DC microgrid model described in step 3 are as follows:

[0113] U min ≤U m ≤U max

[0114] Among them, U m Let U be the node voltage of the m-th node in the DC microgrid model. min U represents the minimum node voltage in the DC microgrid model. max This represents the maximum value of the node voltage in the DC microgrid model;

[0115] The branch current constraints of the DC microgrid model described in step 3 are as follows:

[0116] |I m,n |≤I max

[0117] Among them, I m,n Let I be the branch current between the m-th node and the n-th node in the DC microgrid model. max The current flowing through the branch when the line reaches its maximum heat tolerance;

[0118] The voltage control parameter constraints for the distributed unit of the DC microgrid model described in step 3 are as follows:

[0119] k p,min ≤k p ≤k p,max

[0120] k d,min ≤k d ≤k d,max

[0121] Where, k p k represents the proportional gain of the distributed unit voltage controller in the DC microgrid model. p,min k is the minimum proportional gain of the distributed unit voltage controller. p,max k represents the maximum proportional gain of the distributed unit voltage controller. d Let k be the differential gain of the distributed unit voltage controller in the DC microgrid model. d,min k is the minimum value of the differential gain of the distributed unit voltage controller. d,max This represents the maximum value of the differential gain of the distributed unit voltage controller;

[0122] Step 4: Solve the DC microgrid voltage control optimization model using the improved particle swarm optimization algorithm. Set the initial parameters of the improved particle swarm optimization algorithm, construct the fitness function by combining the eigenvalue parameters of the dynamic matrix of the DC microgrid model, and select the proportional gain and differential gain of the voltage controller of the distributed unit of the DC microgrid model as the decision variables of the optimization algorithm.

[0123] Step 4, which involves setting the initial parameters of the improved particle swarm optimization algorithm, is as follows:

[0124] Initialize the particle search space to have a dimension of 2M = 6, a number of regular particles of D1 = 100, a number of random particles of D2 = 20, and a maximum number of iterations of k. max =500, the inertia weight coefficient of the improved particle swarm algorithm is w, and its value changes randomly within (0,1). The local weight coefficient of the improved particle swarm algorithm is c1=0.5, and the global weight coefficient of the improved particle swarm algorithm is c2=0.5.

[0125] The initial position vector for each particle during the initialization iteration is defined as follows:

[0126]

[0127] j = 1, 2, ..., 2M-1, 2M

[0128] i = 1, 2, ..., D1, D1+1, ..., D1+D2

[0129] in, Let i be the position vector of particle i during the initial iteration, which is the position vector when the iteration number k = 0, and corresponds to the initial proportional gain and initial differential gain of the voltage controller of each distributed unit in the DC microgrid model. Let be the j-th element of the position vector of particle i during the initialization iteration. If j ≤ 3, then it corresponds to the initial proportional gain of the voltage controller of the j-th distributed unit in the DC microgrid model. If j > 3, then it corresponds to the initial differential gain of the voltage controller of the (j-3)-th distributed unit in the DC microgrid model. 2M = 6 is the dimension of the particle search space, which corresponds to the total number of control parameters of the voltage controller of the distributed unit in the DC microgrid. D1 = 100 is the number of regular particles, and D2 = 20 is the number of random particles.

[0130] The elements of the position vector for each particle during the initialization iteration are defined as follows:

[0131]

[0132] j = 1, 2, ..., M

[0133]

[0134] j = M+1, M+2, ..., 2M

[0135] in, Let k be the j-th element of the position vector of particle i during the initial iteration, 2M = 6 be the dimension of the particle search space, rand(0,1) be a random variable in the range (0,1), and k be the position vector of particle i. p,min k is the minimum proportional gain of the distributed unit voltage controller. p,max k represents the maximum proportional gain of the distributed unit voltage controller. d,min k is the minimum value of the differential gain of the distributed unit voltage controller. d,max This represents the maximum value of the differential gain of the distributed unit voltage controller;

[0136] The initial velocity vector for a regular particle during initialization iteration is defined as follows:

[0137]

[0138] j = 1, 2, ..., 2M-1, 2M

[0139] in, Let i be the velocity vector of particle i during the initialization iteration. Let j be the j-th element of the velocity vector of particle i during the initial iteration, and 2M = 6 be the dimension of the particle search space;

[0140] The elements of the velocity vector during the initialization iteration of a regular particle are defined as follows:

[0141]

[0142] j = 1, 2, ..., 2M-1, 2M

[0143] in, Let V be the j-th element of the velocity vector of particle i during the initial iteration, 2M = 6 be the dimension of the particle search space, rand(0,1) be a random variable in the range (0,1), and V be the velocity vector of particle i during the initial iteration. min To improve the minimum particle velocity in the particle swarm optimization algorithm, V max To improve the maximum particle velocity in the particle swarm optimization algorithm;

[0144] The fitness function mentioned in step 4 is:

[0145]

[0146] Where T = 14 is the number of eigenvalues ​​in the dynamic matrix of the DC microgrid model, δ r ω is the real part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r Let a be the imaginary part of the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r b is the damping weighting coefficient for the r-th eigenvalue of the dynamic matrix of the DC microgrid model. r ω is the angular frequency weighting coefficient of the r-th eigenvalue of the dynamic matrix of the DC microgrid model;

[0147] Step 5: Using the position vector elements of each particle at the initial iteration as the initial proportional gain and initial differential gain of the voltage controller of each distributed unit in the DC microgrid model, perform power flow calculation on the DC microgrid model to obtain the voltage of each node and the current of each branch in the DC microgrid model, and update the dynamic matrix and its eigenvalue parameters of the DC microgrid model accordingly. Update the position vector and velocity vector of the regular particles, update the position vector of the random particles, calculate the fitness function value corresponding to the position vector of each particle according to the eigenvalue parameters of the dynamic matrix obtained from the power flow calculation, and update the local optimal solution and global optimal solution of each particle within the constraints.

[0148] Step 5, which involves updating the position and velocity vectors of the regular particles, is as follows:

[0149]

[0150] i = 1, 2, ..., D1

[0151] in, Let i be the position vector of particle i after the k-th iteration. V is the position vector of particle i after the (k+1)th iteration. i k Let V be the velocity vector of particle i after the k-th iteration. i k+1 Let be the velocity vector of particle i after the (k+1)th iteration, rand(0,1) be a random variable in the range (0,1), w be the inertia weight coefficient of the improved particle swarm optimization algorithm, c1 = 0.5 be the local weight coefficient of the improved particle swarm optimization algorithm, and c2 = 0.5 be the global weight coefficient of the improved particle swarm optimization algorithm. Let be the local optimal position vector of particle i after the k-th iteration. Let D1 = 100 be the global optimal position vector after the k-th iteration, where D1 = 100 is the number of regular particles.

[0152] Step 5, updating the position vector of the random particle, is as follows:

[0153]

[0154] i = D1+1, D1+2, ..., D1+D2

[0155] in, Let i be the position vector of particle i after the (k+1)th iteration. All are random variables in the range (0,1) at the k-th iteration, k p,max k represents the maximum proportional gain of the distributed unit voltage controller. d,max D1 = 100 represents the maximum value of the differential gain of the distributed unit voltage controller, and D2 = 20 represents the number of regular particles and random particles.

[0156] Compare the fitness function values ​​f corresponding to the updated position vectors of each particle. i k+1 The fitness function value f of the local optimal solution vector of each particle. i k* Update the local optimal solution vector for each particle:

[0157]

[0158] i = 1, 2, ..., D1 + D2

[0159] in, Let be the local optimal position vector of particle i after the (k+1)th iteration. Let be the position vector of particle i after the (k+1)th iteration. Let be the local optimal position vector of particle i after the k-th iteration. Let f be the fitness function value of the position vector of particle i after k+1 iterations. i k* Let be the fitness function value of the local optimal solution vector of particle i after the k-th iteration;

[0160] The global optimal solution is selected and updated from the local optimal solutions, and its value is as follows:

[0161]

[0162] j = 1, 2, ..., 2M

[0163] g = m i inf i k+1* i = 1, 2, ..., D1 + D2

[0164] in, This is the vector of the globally optimal solution after the (k+1)th iteration. Let f be the j-th element of the global optimal solution vector after the (k+1)-th iteration, g be the number of the particle that minimizes the fitness function value after the (k+1)-th iteration, and f be the element of the vector. i k+1* Let be the fitness function value of the local optimal solution vector after the (k+1)th iteration of the i-th particle, 2M = 6 be the dimension of the particle search space, which corresponds to the total number of control parameters of the voltage controller of the distributed unit of the DC microgrid, D1 = 100 be the number of regular particles, and D2 = 20 be the number of random particles.

[0165] Step 6: Repeat step 5 until the maximum number of iterations is reached to obtain the global optimal solution. Based on the global optimal solution, obtain the optimal proportional gain and optimal differential gain of the voltage controller of each distributed unit in the DC microgrid model to realize real-time control of the DC microgrid voltage.

[0166] The global optimal solution described in step 6 is:

[0167] X g =[X g,1 ,X g,2 ,...X g,j ,...X g,2M ]

[0168] j = 1, 2, ..., 2M

[0169] Among them, X g To utilize the globally optimal solution obtained by the improved particle swarm optimization algorithm, which corresponds to the optimal proportional gain and optimal differential gain of the voltage controller of each distributed unit in the DC microgrid model, X g,j is the j-th element of the global optimal solution vector, and 2M = 6 is the dimension of the particle search space, which corresponds to the total number of control parameters of the voltage controller of the distributed unit of the DC microgrid;

[0170] Step 6 describes obtaining the global optimal solution using the optimization algorithm, which yields the optimal proportional gain and optimal derivative gain of each distributed unit voltage controller, as follows:

[0171]

[0172] l = 1, 2, ..., M

[0173] Where, k p,l Let k be the proportional gain of the voltage controller of the l-th distributed unit in the DC microgrid model. d,l Let X be the differential gain of the voltage controller of the l-th distributed unit in the DC microgrid model. g,l is the l-th element of the global optimal solution vector, and M=3 is the number of distributed units in the DC microgrid model;

[0174] Step 6 describes the implementation of real-time voltage stability control for the DC microgrid, as follows:

[0175] The optimal proportional gain and optimal derivative gain of the voltage controller of each distributed unit are allocated to the generator sets of each distributed unit to achieve real-time voltage control of the DC microgrid.

[0176] A specific embodiment of the present invention also provides a computer-readable medium.

[0177] The computer-readable medium is a server workstation;

[0178] The server workstation stores the computer program executed by the electronic device. When the computer program runs on the electronic device, it causes the electronic device to execute the steps of the vehicle position estimation method of the fusion filtering network according to the present invention.

[0179] It should be understood that any parts not described in detail in this specification belong to the prior art.

[0180] It should be understood that the above description of the preferred embodiments is quite detailed, but it should not be considered as a limitation on the scope of protection of this invention. Those skilled in the art, under the guidance of this invention, can make substitutions or modifications without departing from the scope of protection of the claims of this invention, and all such substitutions or modifications fall within the scope of protection of this invention. The scope of protection of this invention should be determined by the appended claims.

Claims

1. A voltage control method for a DC microgrid, characterized in that, The specific steps are as follows: Step 1: Input the load and line data of the DC microgrid to build a DC microgrid model. The DC microgrid model includes N nodes. Step 2: Construct the state equations for the DC microgrid model. In the DC microgrid model, the first m Each node t The node voltage at time t, In the DC microgrid model, the first n Each node t The node voltage at time t, In the DC microgrid model, the first m Distributed units connected to each node t Output current at any moment For the first m Each node t Load current at any given time In the DC microgrid model, the first m The node and the first n Between nodes t The branch current at time , , , As a state variable, with The state equations of the DC microgrid model are linearized using algebraic variables to obtain the dynamic matrix of the DC microgrid model. Then, the eigenvalues ​​of the dynamic matrix are calculated. in, The first dynamic matrix of the DC microgrid model r 1 eigenvalue, T This represents the number of eigenvalues ​​in the dynamic matrix of the DC microgrid model. δ r The first dynamic matrix of the DC microgrid model r The real part of each eigenvalue The first dynamic matrix of the DC microgrid model r Using the imaginary parts of the eigenvalues, and through eigenvalue analysis of the dynamic matrix of the DC microgrid model, a voltage control optimization model for the DC microgrid model is constructed, defined as follows: in, a r The first dynamic matrix of the DC microgrid model r Damping weight coefficients for each eigenvalue, b r The first dynamic matrix of the DC microgrid model r angular frequency weighting coefficients for each eigenvalue; Step 3: Construct the node voltage constraints, branch current constraints, and voltage control parameter constraints of the distributed units of the DC microgrid model respectively. Step 4: Solve the voltage control optimization model using the improved particle swarm optimization algorithm. Set the initial parameters of the improved particle swarm optimization algorithm, construct the fitness function by combining the eigenvalue parameters of the dynamic matrix of the DC microgrid model, and select the proportional gain and differential gain of the voltage controller of the distributed unit of the DC microgrid model as the decision variables of the optimization algorithm. Step 5: Using the position vector elements of each particle at the initial iteration as the initial proportional gain and initial differential gain of the voltage controller of each distributed unit in the DC microgrid model, perform power flow calculation on the DC microgrid model to obtain the voltage of each node and the current of each branch in the DC microgrid model, and update the dynamic matrix and its eigenvalue parameters of the DC microgrid model accordingly. Update the position vector and velocity vector of the regular particles, update the position vector of the random particles, calculate the fitness function value corresponding to the position vector of each particle according to the eigenvalue parameters of the dynamic matrix obtained from the power flow calculation, and update the local optimal solution and global optimal solution of each particle within the constraints. Step 6: Repeat step 5 until the maximum number of iterations is reached to obtain the global optimal solution. Based on the global optimal solution, obtain the optimal proportional gain and optimal differential gain of the voltage controller of each distributed unit in the DC microgrid model to realize real-time control of the DC microgrid voltage.

2. The DC microgrid voltage control method according to claim 1, characterized in that: Step 1, which describes the construction of a DC microgrid model, involves selecting M nodes from N nodes and incorporating them into a distributed unit.

3. The DC microgrid voltage control method according to claim 2, characterized in that: The state equation of the DC microgrid model described in step 2 is: in, In the DC microgrid model, the first m The output current of the distributed unit connected to each node is in t The derivative at time t, In the DC microgrid model, the first m The node voltage of each node is t The derivative at time t, In the DC microgrid model, the first m The node and the first n The branch current between nodes t The derivative at time t, In the DC microgrid model, the first m The output current of the distributed unit connected to each node is in t The difference between the current time and the previous sampling time. In the DC microgrid model, the first m The node voltage of each node is t The difference between the current time and the previous sampling time. In the DC microgrid model, the first m The node and the first n The branch current between nodes t The difference between the current time and the previous sampling time. m The first DC microgrid model m The node number, n The first DC microgrid model n The node number, N This represents the total number of nodes in the DC microgrid model. t Indicates the first t At that moment, Sampling time, In the DC microgrid model, the first m The equivalent series inductance of the distributed unit connected to each node. In the DC microgrid model, the parallel connection is in the first... m The equivalent capacitance at each node. In the DC microgrid model, the first m The node and the first n Branch inductance between nodes In the DC microgrid model, the first m The node and the first n Branch resistance between nodes E m In the DC microgrid model, the first m The equivalent DC power supply of the distributed unit connected to each node. In the DC microgrid model, the first m The buck chopper of the distributed unit connected to each node is in t Duty cycle at any given moment.

4. The DC microgrid voltage control method according to claim 3, characterized in that: The node voltage constraints of the DC microgrid model described in step 3 are as follows: in, In the DC microgrid model, the first m The node voltage of each node, This represents the minimum node voltage in the DC microgrid model. This represents the maximum value of the node voltage in the DC microgrid model; The branch current constraints of the DC microgrid model described in step 3 are as follows: in, In the DC microgrid model, the first m In the model of the nth node and DC microgrid n Branch current between nodes The current flowing through the branch when the line reaches its maximum heat tolerance; The voltage control parameter constraints for the distributed unit of the DC microgrid model described in step 3 are as follows: in, k p This refers to the proportional gain of the distributed unit voltage controller in the DC microgrid model. k p,min This represents the minimum proportional gain of the distributed unit voltage controller. k p,max This represents the maximum proportional gain of the distributed unit voltage controller. k d This represents the differential gain of the distributed unit voltage controller in the DC microgrid model. k d,min This represents the minimum differential gain of the distributed unit voltage controller. k d,max This represents the maximum value of the differential gain of the distributed unit voltage controller.

5. The DC microgrid voltage control method according to claim 4, characterized in that: Step 4, which involves setting the initial parameters of the improved particle swarm optimization algorithm, is as follows: Initialize the particle search space to have a dimension of 2. M The number of conventional particles is D 1. The number of random particles is D 2. The maximum number of iterations is k max The improved particle swarm optimization algorithm has an inertia weight coefficient of 1. w The improved particle swarm optimization algorithm has the following local weight coefficients: c 1. The global weight coefficient of the improved particle swarm optimization algorithm is: c 2; The initial position vector for each particle during the initialization iteration is defined as follows: in, For particles i The position vector at the initial iteration is the iteration number. k The position vector when =0 corresponds to the initial proportional gain and initial differential gain of the voltage controller of each distributed unit in the DC microgrid model. For particles i The position vector during initialization iteration j If an element, j ≤ M Then, in its corresponding DC microgrid model, the first... j The initial proportional gain of the voltage controller of each distributed unit, if j > M Then, in its corresponding DC microgrid model, the first... j - M The initial differential gain of the distributed unit voltage controller; 2 M The dimension of the particle search space is the total number of control parameters corresponding to the voltage controller of the distributed unit of the DC microgrid. D 1 represents the number of conventional particles. D 2 represents the number of random particles; The elements of the position vector for each particle during the initialization iteration are defined as follows: in, For particles i The position vector during initialization iteration j One element, 2M Let be the dimension of the particle search space. rand (0,1) is a random variable in the range (0,1). k p,min This represents the minimum proportional gain of the distributed unit voltage controller. k p,max This represents the maximum proportional gain of the distributed unit voltage controller. k d,min This represents the minimum differential gain of the distributed unit voltage controller. k d,max This represents the maximum value of the differential gain of the distributed unit voltage controller; The initial velocity vector for a regular particle during initialization iteration is defined as follows: in, For particles i The velocity vector during initialization iteration. For particles i The velocity vector during initialization iteration j 2 elements M Let be the dimension of the particle search space; The elements of the velocity vector during the initialization iteration of a regular particle are defined as follows: in, For particles i The velocity vector during initialization iteration j One element, 2M Let be the dimension of the particle search space. rand (0,1) is a random variable in the range (0,1). V min To improve the minimum particle velocity in the particle swarm optimization algorithm, V max To improve the maximum particle velocity in the particle swarm optimization algorithm.

6. The DC microgrid voltage control method according to claim 5, characterized in that: The fitness function mentioned in step 4 is: in, T The number of eigenvalues ​​in the dynamic matrix of the DC microgrid model. δ r The first dynamic matrix of the DC microgrid model r The real part of each eigenvalue The first dynamic matrix of the DC microgrid model r The imaginary part of each eigenvalue. a r The first dynamic matrix of the DC microgrid model r Damping weight coefficients for each eigenvalue, b r The first dynamic matrix of the DC microgrid model r The angular frequency weighting coefficients of each eigenvalue.

7. The DC microgrid voltage control method according to claim 6, characterized in that: Step 5, which involves updating the position and velocity vectors of the regular particles, is as follows: in, For the first k After the next iteration, the particle i The position vector, For the first k Particle after +1 iteration i The position vector, For the first k After the next iteration, the particle i The velocity vector, For the first k Particle after +1 iteration i The velocity vector, rand (0,1) is a random variable in the range (0,1). w To improve the inertia weight coefficient of the particle swarm optimization algorithm, c 1. To improve the local weight coefficients of the particle swarm optimization algorithm, c 2. To improve the global weight coefficients of the particle swarm optimization algorithm, For the first k After the next iteration, the particle i The local optimal position vector, For the first k The globally optimal position vector after the nth iteration. D 1 represents the number of conventional particles; Step 5, updating the position vector of the random particle, is as follows: in, For the first k Particle after +1 iteration i The position vector, All are the first k The random variable in the range (0,1) during the next iteration k p,max This represents the maximum proportional gain of the distributed unit voltage controller. k d,max This represents the maximum value of the differential gain of the distributed unit voltage controller. D 1 represents the number of conventional particles. D 2 represents the number of random particles; Compare the fitness function values ​​corresponding to the updated position vectors of each particle. The fitness function value of the local optimal solution vector of each particle. Update the local optimal solution vector for each particle: in, For particles i In the k The local optimal position vector after +1 iterations For particles i In the k The position vector after +1 iterations For particles i In the k The local optimal position vector after the nth iteration. For particles i exist k The fitness function value of the position vector after +1 iterations. For particles i No. k The fitness function value of the local optimal solution vector after the next iteration; The global optimal solution is selected and updated from the local optimal solutions, and its value is as follows: in, For the first k The global optimal solution vector after +1 iterations For the first k After +1 iterations, the global optimal solution vector is... j One element, g For the first k The number of the particle that minimizes the fitness function after +1 iterations. For the first i Particles k The fitness function value of the local optimal solution vector after +1 iterations, 2 M The dimension of the particle search space corresponds to the total number of control parameters of the voltage controller in the distributed unit of the DC microgrid. D 1 represents the number of conventional particles. D 2 represents the number of random particles.

8. The DC microgrid voltage control method according to claim 7, characterized in that: The global optimal solution described in step 6 is: in, To utilize the globally optimal solution obtained by the improved particle swarm optimization algorithm, which corresponds to the optimal proportional gain and optimal differential gain of the voltage controller of each distributed unit in the DC microgrid model, The first of the global optimal solution vectors j 2 elements M The dimension of the particle search space is the total number of control parameters corresponding to the voltage controller of the distributed unit of the DC microgrid. Step 6 describes obtaining the global optimal solution using the optimization algorithm, which yields the optimal proportional gain and optimal derivative gain of each distributed unit voltage controller, as follows: in, In the DC microgrid model, the first l The proportional gain of the distributed unit voltage controller In the DC microgrid model, the first l The differential gain of a distributed unit voltage controller The first of the global optimal solution vectors l One element, M This represents the number of distributed units in the DC microgrid model. Step 6 describes the implementation of real-time voltage stability control for the DC microgrid, as follows: The optimal proportional gain and optimal derivative gain of the voltage controller of each distributed unit are allocated to the generator sets of each distributed unit to achieve real-time voltage control of the DC microgrid.

9. A computer-readable medium, characterized in that, It stores a computer program executed by an electronic device, which, when run on the electronic device, causes the electronic device to perform the steps of the method as described in any one of claims 1-8.

Images