Power distribution network voltage collaborative optimization method based on improved artificial travel mouse algorithm

Abstract

The invention discloses a power distribution network voltage collaborative optimization method based on an improved artificial travel mouse algorithm, and the method comprises the following steps: firstly, building a power distribution network physical model based on an IEEE33 node topological graph and the positions of three voltage regulation devices, namely a capacitor, a distributed power supply and an energy storage device; then, aiming at the physical model of the power distribution network, establishing a comprehensive objective function of voltage deviation, network loss and full life cycle cost, and setting constraint conditions according to the characteristics of the power distribution network and the three voltage regulating devices; and finally, optimizing and iterating by adopting an improved artificial travel mouse optimization algorithm to obtain the optimal output power values of the three kinds of voltage regulating equipment. According to the method, Logistic chaotic mapping and Gaussian variation are added into the artificial Louver-mouse algorithm, so that the generated initial position has better ergodicity and unpredictability, the search process has better randomness and diversity, the voltage deviation and network loss of the power distribution network are effectively reduced, and the stability of the power distribution network is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of distribution network voltage control, and particularly relates to a distribution network voltage collaborative optimization method based on an improved artificial vole algorithm. BACKGROUND

[0002] Under the background of current energy transformation acceleration, large-scale renewable energy grid connection, electric vehicle charging load surge and wide access of flexible load, the distribution network voltage fluctuation problem is increasingly prominent. Intermittent wind power output causes voltage sag and temporary rise, photovoltaic power fluctuation causes frequency deviation, and concentrated electric vehicle charging brings harmonic pollution and aggravates peak-valley difference. The existing voltage regulation mainly relies on independent control of distributed power, capacitor or energy storage device, and there are problems such as lack of coordination between devices, insufficient regulation accuracy, slow response speed and the like, which are difficult to cope with complex working conditions under high proportion of new energy access. Therefore, the present application provides a distribution network voltage collaborative optimization method based on an improved artificial vole algorithm, which uses the improved artificial vole optimization algorithm to find the optimal output power value of different devices, effectively reduces the voltage deviation and network loss of the system, and significantly improves the stability of the distribution network. SUMMARY

[0003] To achieve the above-mentioned purpose, the technical scheme provided by the present application is as follows:

[0004] S1: based on the IEEE33 node topology graph and the positions of the three types of voltage regulating devices, capacitor, distributed power and energy storage device, a distribution network physical model is established;

[0005] S1-1: based on the topology structure of the standard IEEE33 node, the connection relationship of 33 nodes and 32 branches is defined, the node data includes the active load and reactive load of each node, and the branch data includes the starting node, the ending node, the branch resistance and the branch reactance of each line;

[0006] S1-2: the forward-backward substitution method is used for power flow calculation, and a radial network structure is established through the branch connection matrix, wherein node 1 is taken as the balance node;

[0007] S1-3: based on the IEEE33 node topology, the installation positions of the three types of voltage regulating devices are determined, the capacitor is configured at the weak voltage nodes [6, 15, 24], the distributed power is deployed at the end nodes [18, 25, 30], and the energy storage device is installed at the key load nodes [7, 22, 31];

[0008] S2: for the distribution network physical model, a comprehensive objective function of voltage deviation, network loss and full life cycle cost is established, and constraint conditions are set according to the characteristics of the distribution network and the three types of voltage regulating devices;

[0009] S2-1: the comprehensive objective function is established:

[0010] MinF = ω1·f volt + ω2·f loss + ω3·f cost (1)

[0011] In the formula, F is the comprehensive target function, f volt is the voltage deviation, f loss is the network loss, f cost is the full life cycle cost, ω1-ω3 are weight coefficients, which are 0.3, 0.5, 0.2 respectively;

[0012] S2-2: the voltage deviation f volt is:

[0013]

[0014] In the formula, n is the total number of nodes, i is the current node number, U i is the voltage amplitude of node i, U i ' is the voltage expectation value of node i, U imax , U imin are the upper and lower limits of the voltage value of node i respectively;

[0015] S2-3: the network loss f loss is:

[0016]

[0017] In the formula, c is the current branch number, B is the total number of branches, I c is the current amplitude of the cth branch, R c is the resistance value of the cth branch;

[0018] S2-4: the full life cycle cost f cost is divided into investment cost C inv and operation and maintenance cost C OM , and its formula is:

[0019]

[0020] In the formula, r is the discount rate; T is the project life; n cap , n pv , n Ess are the access quantities of capacitors, distributed power sources and energy storage devices respectively; Q inv , P inv , S inv are the capacities of capacitors, distributed power sources and energy storage devices respectively; Q OM , P OM , S OMThe output power of the capacitor, the distributed power supply, and the energy storage device; x, y, z are the unit price of the capacity of the capacitor, the distributed power supply, and the energy storage device respectively; H, M, Y are the annual average operation and maintenance cost of the capacitor, the distributed power supply, and the energy storage device respectively; d, e, f are the current numbers of the capacitor, the distributed power supply, and the energy storage device respectively;

[0021] S2-5: The complete formula of the comprehensive objective function is obtained as follows:

[0022]

[0023] S2-6: The power flow constraint contains the active power P i and the reactive power Q i injected into the current node i, and the formula is as follows:

[0024]

[0025] In the formula, U j is the voltage amplitude of the node j, θ ij is the angle difference between the voltages at the two ends of the node, G ij and B ij are the conductance value and the susceptance value between the node i and the node j on the line respectively, sin is the sine function, cos is the cosine function, and j represents the imaginary unit;

[0026] S2-7: The line allowed active power and reactive power constraint is as follows:

[0027]

[0028] In the formula, P Lmax and Q Lmax are the maximum active power and the maximum reactive power allowed to flow through the line respectively;

[0029] S2-8: The capacitor constraint is as follows:

[0030] 0≤Q Ci ≤Q Cmax (10)

[0031] In the formula, Q Ci is the reactive power compensation of the node i, and Q Cmax is the maximum reactive power compensation of the capacitor;

[0032] S2-9: The distributed power supply constraint is as follows:

[0033] P DGi.min ≤P DGi ≤P DGi.max (11)

[0034] In the formula, P DGiP represents the active power of the energy storage device at node i. DGi.max P DGi.min These represent the maximum and minimum active power of the energy storage device, respectively.

[0035] S2-10: Constraints for energy storage devices are:

[0036] P Essi.min ≤P Essi ≤P Essi.max (12)

[0037] In the formula, P Essi P represents the current active power of the energy storage device at node i. Essi.max P Essi.min These represent the maximum and minimum active power of the energy storage device, respectively.

[0038] S3: An improved artificial lemming optimization algorithm is used for iterative optimization to obtain the optimal output power values ​​for the three voltage regulating devices;

[0039] S3-1: The artificial lemming optimization algorithm is based on a population, treating each lemming as a candidate solution, and using the output power values ​​of capacitors, distributed power sources, and energy storage devices as lemming position parameters. The total number of lemmings is the population size N.

[0040] S3-2: Use Logistic chaotic mapping to generate the initial positions of all lemmings:

[0041] x i+1 =wx i (1-x i (13)

[0042] In the formula, x i+1 For the individual at the next moment; x i The current individual; w is a random number between (3,4);

[0043] S3-3: For each lemming, the energy factor E(t) determines whether it enters the exploration phase or the development phase:

[0044]

[0045] In the formula, t is the current iteration number, T max t represents the maximum number of iterations, rand is a random number, arctan is the arctangent function, and ln is the logarithmic function. When the energy factor E(t) > 1, the exploration phase begins; when the energy factor E(t) ≤ 1, the development phase begins. As the number of iterations t increases, the probability of entering the development phase also increases.

[0046] S3-4: During the exploration phase, the location of lemming updates is determined by the lemmings' migration and burrowing behaviors, and the specific update method is as follows:

[0047]

[0048] where Z s (t+1) is the position of the s-th search agent at the t+1 iteration, Z s (t) is the current position of the s-th search agent, Z best (t) is the current best solution, Z a (t) is the individual randomly selected when rand>0.3, Z b (t) is the individual randomly selected when rand≤0.3, i.e., other, f BM (x;0,1) is a normal distribution with variance 1 and mean 0 under standard Brownian motion, R is a vector with size 1 x Dim, Dim is the total number of three kinds of devices, i.e., capacitor, distributed power supply and energy storage device, L is a random number related to the current iteration number, L and Z b (t) is used to describe the interaction of the vole individuals when digging new holes; f BM The specific formulas of R, L and L are as follows:

[0049]

[0050] R = 2 x rand(1, Dim)-1 (17)

[0051]

[0052] where x b represents a random variable of standard normal distribution, exp is an exponential function, and π is a circular constant;

[0053] S3-5: In the development stage, the updated position of the vole is determined by the foraging and predator avoidance behaviors of the vole, and the specific updating method is as follows:

[0054]

[0055] where spiral is a spiral shape of random search in the foraging process, G is the escape coefficient of the vole, Levy represents the Levy flight function, and Levy(Dim) represents a Dim-dimensional Levy random vector, wherein each dimension is an independent Levy random number, which is used to simulate the deceptive action of the vole when escaping; the formulas of spiral, G and Levy are as follows:

[0056]

[0057]

[0058] In the formula, u and v are random values in the interval [0-1], sigma1 is a scale parameter, beta is 1.5, Gamma is the gamma function, and h is the current number of devices;

[0059] S3-6: Add Gaussian variation disturbance in the development stage to make the algorithm jump out of the local optimum, and the formula is:

[0060]

[0061] In the formula, alpha is a random number in [0, 1], and sigma2 is a random number in (0, 1];

[0062] S3-7: Improve the fixed value sigma2 to have an adaptive strategy, and the improved formula is:

[0063] sigma2=0.1·e -λt (25)

[0064] In the formula, lambda is the decay rate, lambda=1 / T max ;

[0065] After Gaussian variation disturbance, the position is updated to the disturbed position, and the formula is:

[0066] x g+1 =x g +G(alpha) (26)

[0067] In the formula, x g+1 is the algorithm input value after Gaussian variation, and x g is the algorithm input value before Gaussian variation. In order to ensure that the disturbed position is within the set value range, the boundary condition is set:

[0068]

[0069] In the formula, u b is the upper limit of the set value, and l b is the lower limit of the set value. After boundary processing, the updated value can be ensured to be within the set range;

[0070] S3-8: Calculate the fitness value of the updated vole position of steps S3-2 and S3-7, and sort them according to their size to find the optimal result of the current iteration;

[0071] S3-9: Repeat steps S3-3 to S3-8 until the maximum number of iterations is completed, and the vole position with the smallest fitness is obtained, which is the optimal power allocation value.

[0072] Compared with the prior art, the principles and advantages of the present scheme are as follows:

[0073] The application discloses a power distribution network voltage collaborative optimization method based on an improved artificial lemming algorithm, first, based on the IEEE33 node topology graph and the positions of three voltage regulating devices, namely capacitors, distributed power sources and energy storage devices, a power distribution network physical model is established; then, for the power distribution network physical model, a comprehensive objective function of voltage deviation, network loss and full life cycle cost is established, and constraint conditions are set according to the characteristics of the power distribution network and the three voltage regulating devices; finally, the improved artificial lemming optimization algorithm is used for optimization iteration, and the optimal output power values of the three voltage regulating devices are obtained. The application adds Logistic chaotic mapping and Gaussian mutation in the artificial lemming algorithm, so that the generated initial positions have better ergodicity and unpredictability, the search process has better randomness and diversity, the voltage deviation and network loss of the power distribution network are effectively reduced, and the stability of the power distribution network is significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0074] Figure 1 A flowchart of the power distribution network voltage collaborative optimization method based on the improved artificial lemming algorithm in the embodiment of the application;

[0075] Figure 2 An IEEE33 node topology structure diagram after different devices are accessed in the embodiment of the application;

[0076] Figure 3 A voltage amplitude diagram after the artificial lemming algorithm optimization in the embodiment of the application;

[0077] Figure 4 A network loss diagram after the artificial lemming algorithm optimization in the embodiment of the application;

[0078] Figure 5 A cost comparison diagram after the artificial lemming algorithm optimization in the embodiment of the application;

[0079] Figure 6 A device power value output diagram after the artificial lemming algorithm optimization in the embodiment of the application. DETAILED DESCRIPTION

[0080] The application will be further described below in combination with specific embodiments:

[0081] Figure 1 A flowchart of the power distribution network voltage collaborative optimization method based on the improved artificial lemming algorithm, Figure 2 An IEEE33 node topology structure diagram after different devices are accessed; in the system, there are three devices, namely capacitors, distributed power sources and energy storage devices, and each device is connected to three devices, and there are nine devices in total, and the specific optimization process includes the following steps:

[0082] S1: based on the IEEE33 node topology graph and the positions of three voltage regulating devices, namely capacitors, distributed power sources and energy storage devices, a power distribution network physical model is established;

[0083] S1-1: Based on the topology structure of standard IEEE33 node, the connection relationship of 33 nodes and 32 branches is defined, the node data includes the active load and reactive load of each node, and the branch data includes the starting node, terminal node, branch resistance and branch reactance of each line;

[0084] S1-2: The forward-backward substitution method is used for power flow calculation, and the radial network structure is established through the branch connection matrix, wherein node 1 is taken as the balance node;

[0085] S1-3: On the basis of IEEE33 node topology, the installation positions of three types of voltage regulating devices are determined, the capacitor is configured at the weak voltage node [6, 15, 24], the distributed power supply is deployed at the end node [18, 25, 30], and the energy storage device is installed at the key load node [7, 22, 31];

[0086] S2: For the physical model of distribution network, a comprehensive objective function of voltage deviation, network loss and full life cycle cost is established, and according to the characteristics of distribution network and three types of voltage regulating devices, the constraint conditions are set;

[0087] S2-1: The comprehensive objective function is established as follows:

[0088] Min F = ω1·f volt + ω2·f loss + ω3·f cost (28)

[0089] In the formula, F is the comprehensive objective function, f volt is the voltage deviation, f loss is the network loss, f cost is the full life cycle cost, ω1-ω3 are weight coefficients, which are 0.3, 0.5 and 0.2 respectively;

[0090] S2-2: The voltage deviation f volt is:

[0091]

[0092] In the formula, n is the total number of nodes, i is the current node number, U i is the voltage amplitude of node i, U i ’ is the voltage expected value of node i, U imax , U imin are the upper limit and lower limit of the voltage value of node i respectively;

[0093] S2-3: The network loss f loss is:

[0094]

[0095] where c is the current branch number, B is the total number of branches, I c is the current amplitude of the cth branch, R c is the resistance value of the cth branch.

[0096] S2-4: Full life cycle cost f cost is divided into investment cost C inv and operation and maintenance cost C OM , whose formula is:

[0097]

[0098]

[0099] where r is the discount rate; T is the project life; n cap , n pv , n Ess are the number of capacitor, distributed power supply, and energy storage device access, respectively; Q inv , P inv , S inv are the capacity of capacitor, distributed power supply, and energy storage device, respectively; Q OM , P OM , S OM are the output power of capacitor, distributed power supply, and energy storage device; x, y, z are the unit price of capacitor, distributed power supply, and energy storage device, respectively; H, M, Y are the annual average operation and maintenance cost of capacitor, distributed power supply, and energy storage device, respectively; d, e, f are the current number of capacitor, distributed power supply, and energy storage device, respectively.

[0100] S2-5: The complete formula of the comprehensive objective function is:

[0101]

[0102] S2-6: The power flow constraint contains the active power P i and the reactive power Q i injected into the current node i, whose formula is:

[0103]

[0104] where U j is the voltage amplitude of node j, θ ij is the angle difference between the voltages at the beginning and end of the node, G ij , B ij are the conductance and susceptance between node i and node j on the line, respectively, sin is the sine function, cos is the cosine function, and j represents the imaginary unit.

[0105] S2-7: The line allows active power and reactive power constraints are:

[0106]

[0107] In the formula, P Lmax , Q Lmax The maximum active power and maximum reactive power allowed to flow through the line, respectively;

[0108] S2-8: Capacitor constraints are:

[0109] 0≤Q Ci ≤Q Cmax (37)

[0110] In the formula, Q Ci The reactive power compensation of node i, Q Cmax The maximum reactive power compensation of the capacitor;

[0111] S2-9: Distributed power constraints are:

[0112] P DGi.min ≤P DGi ≤P DGi.max (38)

[0113] In the formula, P DGi The active power of energy storage device at node i, P DGi.max , P DGi.min The maximum and minimum active power of the energy storage device, respectively;

[0114] S2-10: Energy storage device constraints are:

[0115] P Essi.min ≤P Essi ≤P Essi.max (39)

[0116] In the formula, P Essi The current active power of the energy storage device at node i, P Essi.max , P Essi.min The maximum and minimum active power of the energy storage device, respectively;

[0117] S3: The improved artificial vole optimization algorithm is used for optimization iteration to obtain the optimal output power value of the three voltage regulating devices;

[0118] S3-1: The artificial vole optimization algorithm is based on population, each vole is regarded as a candidate solution, and the output power value of the capacitor, distributed power and energy storage device is regarded as the vole position parameter. The total number of voles is the population size N.

[0119] S3-2: Use Logistic chaotic mapping to generate the initial position of all voles:

[0120] x i+1 = wx i (1-x i ) (40)

[0121] where x i+1 is the individual at next time; x i is the individual at current time; w is a random number between (3, 4];

[0122] S3-3: For each lemming, decide whether it enters the exploration phase or the exploitation phase according to the energy factor E(t):

[0123]

[0124] where t is the current iteration number, T max is the maximum iteration number, rand is a random number, arctan is the inverse tangent function, and ln is the logarithmic function; when the energy factor E(t) > 1, it enters the exploration phase, and when the energy factor E(t) ≤ 1, it enters the exploitation phase; as the iteration number t increases, the probability of entering the exploitation phase increases;

[0125] S3-4: In the exploration phase, the updated position of the lemming is determined by the migration and hole-digging behavior of the lemming, and the specific updating method is as follows:

[0126]

[0127] where Z s (t+1) is the position of the s-th search agent at the t+1 iteration, Z s (t) is the current position of the s-th search agent, Z best (t) is the current optimal solution, Z a (t) is a randomly selected individual when rand > 0.3, Z b (t) is a randomly selected individual when rand ≤ 0.3, i.e., other, f BM (x; 0, 1) is a normal distribution with variance 1 and mean 0 under standard Brownian motion, R is a vector with size 1 x Dim, Dim is the total number of capacitors, distributed power sources, and energy storage devices, L is a random number related to the current iteration number, and L and Z b (t) are used to describe the interaction of lemming individuals when digging new holes; the specific formulas of f BM , R, and L are as follows:

[0128]

[0129] R = 2 x rand(1, Dim) - 1 (44)

[0130]

[0131] In the formula, x b Let exp be a random variable representing a standard normal distribution, π be an exponential function, and π be the mathematical constant pi.

[0132] S3-5: During the development phase, the location of lemming updates is determined by the lemmings' foraging and predator avoidance behaviors. The specific update method is as follows:

[0133]

[0134] In the formula, *spiral* represents the spiral shape of the random search during foraging, *G* is the lemming's escape coefficient, *Levy* represents the Levy flight function, and *Levy(Dim)* represents generating a *Dim*-dimensional Levy random vector, where each dimension is an independent Levy random number used to simulate the deceptive behavior of the lemming during escape; the formulas for *spiral*, *G*, and *Levy* are:

[0135]

[0136] In the formula, u and ν are random values ​​in the interval [0-1], σ1 is the scale parameter, β is 1.5, Γ is the gamma function, and h is the number of current devices;

[0137] S3-6: During the development phase, a Gaussian mutation perturbation is added to help the algorithm escape local optima. The formula is as follows:

[0138]

[0139] In the formula, α is a random number in [0,1], and σ2 is a random number in (0,1].

[0140] S3-7: An improvement is made to the constant σ2 to give it an adaptive strategy. The improved formula is as follows:

[0141] σ² = 0.1·e -λt (52)

[0142] In the formula, λ is the decay rate, λ=1 / T max ;

[0143] After Gaussian mutation perturbation, the position is updated to the perturbed position, and the formula is:

[0144] x g+1 =x g +G(α) (53)

[0145] In the formula, x g+1 x is the algorithm input value after Gaussian mutation. gThe algorithm input value before Gaussian variation is set as a boundary condition constraint to ensure that the position after variation is within the set value range:

[0146]

[0147] In the formula, u b is the upper limit of the set value, and l b is the lower limit of the set value. After boundary processing, it can be ensured that the updated value is within the set range.

[0148] S3-8: Calculate the fitness value of the lemming position updated in steps S3-2 and S3-7, and sort them according to their sizes to find the optimal result of the current iteration.

[0149] S3-9: Repeat steps S3-3 to S3-8 until the maximum number of iterations set is completed, and obtain the lemming position with the minimum fitness, which is the optimal power allocation value.

[0150] Figure 3 The voltage amplitude diagram after optimization by the improved artificial lemming algorithm is shown. The minimum voltage after the artificial lemming optimization algorithm is 0.942pu, which is 0.026pu higher than the original voltage. The minimum voltage after the improved artificial lemming optimization algorithm is 0.953pu, which is 0.037pu higher than the original voltage.

[0151] Figure 4 The network loss diagram after optimization by the improved artificial lemming algorithm is shown. The network loss value after the artificial lemming optimization algorithm is 0.1055MW, which is 0.0972MW lower than the original network loss. The network loss value after the improved artificial lemming optimization algorithm is 0.0894MW, which is 0.1133MW lower than the original network loss.

[0152] Figure 5 The cost comparison diagram after optimization by the improved artificial lemming algorithm is shown. The operation and maintenance cost after the artificial lemming optimization algorithm is 5.677×10 3 $, which is 2.81×10 3 $ higher than the original operation and maintenance cost. The full life cycle cost is 397.783×10 3 $, which is 27.54×10 3 $ higher than the original full life cycle cost. The operation and maintenance cost after the improved artificial lemming optimization algorithm is 8.16×10 3 $, which is 5.29×10 3 $ higher than the original operation and maintenance cost. The full life cycle cost is 422.116×10 3 $, which is 51.92×10 3 $ higher than the original full life cycle cost.

[0153] Figure 6 The device power output value chart before and after improvement is shown, capacitors, distributed power supply, energy storage device, three of each, a total of nine devices, the initial power output value of the device is [0.25MV, 0.2MV, 0.3MV, 0.08MV, 0.06MV, 0.1MV, 0.05MV, 0.06MV, 0.04MV], the output power of each device after artificial lemming algorithm optimization is [0.08MV, 0.62MV, 0.08MV, 0.2MV, 0.03MV, 0.2MV, 0.15MV, 0.02MV, 0.15MV], the output power of each device after improved artificial lemming algorithm optimization is [0.8MV, 0.8MV, 0.8MV, 0.2MV, 0.2MV, 0.2MV, 0.15MV, 0.15MV, 0.15MV], the power output value of the capacitor with the lowest construction unit price after optimization is obviously increased, compared with the original power output value, the highest increase is 0.42MV, 0.3MV respectively under the artificial lemming algorithm and the improved artificial lemming algorithm.

[0154] From the above analysis, it can be seen that the power distribution network voltage cooperative optimization method based on the improved artificial lemming algorithm effectively reduces the voltage deviation and network loss of the system, and significantly improves the stability of the power distribution network.

[0155] The above-mentioned embodiments are only preferred embodiments of the present application, and do not limit the scope of the application, so any changes made according to the shape and principle of the present application should be covered within the protection scope of the present application.

Claims

1. A power distribution network voltage collaborative optimization method based on an improved artificial vole algorithm, characterized in that, The method comprises the following steps: S1: based on IEEE33 node topology and capacitor, distributed power supply, energy storage device three kinds of voltage regulating equipment position, establish distribution network physical model; S1-1: based on the topology structure of standard IEEE33 node, define the connection relationship of 33 nodes and 32 branches, node data includes the active load and reactive load of each node, branch data includes the starting node, terminal node, branch resistance and branch reactance of each line; S1-2: adopt back and forth method for power flow calculation, establish radial network structure through branch connection matrix, wherein node 1 is taken as balance node; S1-3: on the basis of IEEE33 node topology, the installation position of three kinds of voltage regulating equipment is clear, capacitor is configured at weak voltage node [6, 15, 24], distributed power supply is deployed at terminal node [18, 25, 30], and energy storage device is installed at key load node [7, 22, 31]; S2: for the distribution network physical model, a comprehensive objective function of voltage deviation, network loss and life cycle cost is established, and constraint conditions are set according to the characteristics of distribution network and three kinds of voltage regulating equipment; S2-1: the comprehensive objective function is established: MinF = ω1 · f volt + ω2 · f loss + ω3 · f cost (1) In the formula, F is a comprehensive target function, f volt is a voltage deviation, f loss is a network loss, f cost is a full life cycle cost, ω1-ω3 are weight coefficients, and are 0.3, 0.5, and 0.2, respectively. S2-2: Voltage deviation f volt is: where n is the total number of nodes, i is the current node number, U i is the voltage amplitude of node i, U i is the voltage expectation value of node i, U imax , U imin are the upper and lower limits of the voltage value of node i, respectively S2-3: network loss f loss is: where c is the current branch number, B is the total number of branches, I c is the current amplitude of the cth branch, R c is the resistance value of the cth branch; S2-4: Life cycle cost f cost is divided into investment cost C inv and operation and maintenance cost C OM , and the formula is: In the formula, r is the discount rate; T is the project life; n cap , n pv , n Ess respectively are the access quantity of capacitors, distributed power sources and energy storage devices; Q inv , P inv , S inv respectively are the capacity of capacitors, distributed power sources and energy storage devices; Q OM , P OM , S OM are the output power of capacitors, distributed power sources and energy storage devices; x, y, z respectively are the unit price of the capacity of capacitors, distributed power sources and energy storage devices; H, M, Y respectively are the annual average operation and maintenance cost of capacitors, distributed power sources and energy storage devices; d, e, f respectively are the current number of capacitors, distributed power sources and energy storage devices; S2-5: the complete formula of the comprehensive objective function is: S2-6: The current flow constraint includes the active power P injected into the current node i i and the reactive power Q i The formula is: wherein U j is the voltage amplitude of node j, θ ij is the angular difference of the voltage between the first and the last node of the node, G ij , B ij are the conductance value and the susceptance value between node i and node j on the line, respectively, sin is the sine function, cos is the cosine function, and j represents the imaginary unit; S2-7: the line allowed active power and reactive power constraint is: In the formula, P Lmax , Q Lmax are the maximum active power and the maximum reactive power allowed to flow through the line, respectively. S2-8: the capacitor constraint is: 0 ≤ Q Ci ≤ Q Cmax (10) In the formula, Q Ci Qi is the reactive power compensation amount of node i, Q Cmax Qi is the maximum reactive power compensation amount of the capacitor S2-9: the distributed power supply constraint is: P DGi.min ≤P DGi ≤P DGi.max (11) In the formula, P DGi is the active power of the energy storage device when the node i DGi.max is the maximum value of the active power of the energy storage device, and P DGi.min is the minimum value of the active power of the energy storage device. S2-10: the energy storage device constraint is: P Essi.min ≤P Essi ≤P Essi.max (12) In the formula, P Essi is the current active power of the energy storage device when the node i Essi.max , P Essi.min are the maximum and minimum values of the active power of the energy storage device, respectively. S3: the optimal output power value of three kinds of voltage regulating equipment is obtained by using improved artificial vole optimization algorithm for optimization iteration; S3-1: the artificial vole optimization algorithm is based on population, each vole is taken as a candidate solution, and the output power value of capacitor, distributed power supply and energy storage device is taken as vole position parameter, and the total number of voles is the population size N; S3-2: the initial position of all voles is generated by using Logistic chaos mapping: x i+1 = wx i (1-x i ) (13) In the formula, x i+1 is the individual at the next time; x i is the individual at the current time w is a random number between (3, 4]; S3-3: for each vole, whether it enters the exploration stage or the development stage is determined according to the energy factor E(t): where t is the current iteration number, T max is the maximum iteration number, rand is a random number, arctan is the inverse tangent function, and ln is the logarithm function; when the energy factor E(t) > 1, the exploration phase is entered, and when the energy factor E(t) ≤ 1, the development phase is entered; as the iteration number t increases, the probability of entering the development phase increases; S3-4: in the exploration stage, the updated position of the vole is determined by the migration and hole digging behavior of the vole, and the specific updating mode is as follows: where Z s (t+1) is the position of the s-th search agent at the t+1 iteration, Z s (t) is the current position of the s-th search agent, Z best (t) is the current optimal solution, Z a (t) is the individual randomly selected when rand>0.3, Z b (t) is the individual randomly selected when rand≤0.3, i.e., other, f BM (x;0,1) is a normal distribution with variance 1 and mean 0 under standard Brownian motion, R is a vector with size 1xDim, Dim is the total number of three kinds of devices, i.e., capacitor, distributed power supply and energy storage device, L is a random number related to the current iteration number, L and Z b (t) is used to describe the interaction of the vole individuals when digging new holes; f BM The specific formulas of R, L and Z R = 2 x rand (1, Dim) - 1 (17) where x b represents a random variable of standard normal distribution, exp is an exponential function, and π is a circular constant. S3-5: in the development stage, the updated position of the vole is determined by the foraging and predator avoidance behavior of the vole, and the specific updating mode is as follows: In the formula, spiral is the spiral shape of random search in the foraging process, G is the escape coefficient of the vole, Levy represents the Levy flight function, Levy (Dim) represents a Dim-dimensional Levy random vector, wherein each dimension is an independent Levy random number, which is used to simulate the deception action of the vole when escaping; the formula of spiral, G and Levy is: In the formula, u and v are random values in the interval [0-1], σ1 is a scale parameter, β is 1.5, Γ is the gamma function, and h is the current device number; S3-6: in the development stage, Gaussian variation disturbance is added to make the algorithm jump out of the local optimum, and the formula is: In the formula, α is a random number in [0, 1], and σ2 is a random number in (0, 1]; S3-7: The value σ2 is improved to have an adaptive strategy, and the improved formula is: σ2= 0.1 e -λt (25) where λ is the decay rate, λ = 1 / T max ; After the Gaussian mutation disturbance, the position is updated to the disturbed position, and the formula is: x g+1 = x g + G(a) (26) In the formula, x g+1 is the algorithm input value after Gaussian mutation, x g is the algorithm input value before Gaussian mutation. In order to ensure that the mutated position is within the set value range, a boundary condition constraint is set: wherein u b is an upper limit of the set value, and b is a lower limit of the set value, and the boundary processing is performed to ensure that the updated value is within the set range. S3-8: The fitness values of the positions of the voles updated in steps S3-2 and S3-7 are calculated, and are sorted according to the sizes to find the optimal result of the current iteration; S3-9: Steps S3-3 to S3-8 are repeated until the set maximum number of iterations is completed, and the position of the vole with the minimum fitness is obtained, that is, the optimal power allocation value.

Images