A comprehensive resource optimization configuration method after large-scale application of decentralized electric heating

Abstract

A comprehensive resource optimization configuration method after the large-scale application of decentralized electric heating, including several links such as establishing a two-layer power grid comprehensive resource optimization configuration model, using the particle swarm algorithm to find the optimal solution of the configuration model, using the genetic algorithm to obtain the optimal result of the scheduling model, and obtaining the optimal configuration plan of the comprehensive resources of the regional distribution network; this two-layer planning model considers the practicability and economy of installing power quality control devices, installs the TSC+APF device near the low-voltage distribution system side of decentralized electric heating users to achieve distributed control. In addition, on the premise of considering both economy and power quality control effect, the optimization configuration of the number and location of phase-change switches is modeled, and the optimization adjustment of the load is considered to control the three-phase imbalance problem of the substation area.

Description

Technical Field

[0001] The present invention relates to the technical field of power quality analysis and measure governance after the large-scale application of distributed electric heating loads, and particularly relates to a comprehensive resource optimization configuration method after the large-scale application of distributed electric heating.

Background Art

[0002] With the increasing severity of environmental pollution and energy crisis problems, the country has paid more and more attention to energy conservation and emission reduction and optimizing the energy consumption structure. In view of the current severe situation of environmental change, China put forward at the United Nations General Assembly held on September 22, 2020, that it strives to reach the peak of carbon dioxide emissions before 2030 and aims to achieve carbon neutrality before 2060. Under the dual pressure of resources and environment, it is urgent to find new heating alternative methods to improve the clean utilization rate of energy and alleviate the current development contradictions. At present, most heat sources use coal-fired boilers with traditional fossil energy, which have great advantages in terms of economy. However, the heating method using coal-fired boilers will exacerbate the growth of carbon emissions, and the advantages of coal-fired boilers in terms of energy efficiency are not obvious. Air source heat pumps have high electro-thermal conversion efficiency and are the main "coal-to-electricity" heating power substitution equipment at present. The large-scale access of electric heating equipment has brought great pressure to the stable operation of the distribution network. For example, when a large-scale electric heating load is connected to the rural distribution network, there are problems such as insufficient transformer capacity, small wire current-carrying capacity, and uneven load distribution in the original rural low-voltage distribution network, which are difficult to meet the electricity demand after the "coal-to-electricity" project, and the economic reliability of the system operation is greatly affected. It is necessary to re-plan, design, and expand the rural distribution network; in Huaibei area of Anhui Province, electric heating has been vigorously promoted, resulting in a serious low supply voltage in some low-voltage areas during the peak electricity consumption period; in Le'an County of Jiangxi Province, serious three-phase electricity imbalance has occurred due to the promotion of the electric heating project.

[0003] At present, most of the research on the optimization of distribution network resources at home and abroad focuses on urban distribution networks and rural distribution networks, while ignoring the optimization of distribution network resources after the large-scale application of distributed electric heating. The power quality problems of the distribution network after the large-scale application of distributed electric heating are more serious and the resource optimization is more complex. This patent proposes a comprehensive resource optimization configuration method for the large-scale application of distributed electric heating.

Summary of the Invention

[0004] The purpose of the present invention is to provide a comprehensive resource optimization configuration method after the large-scale application of distributed electric heating. It can overcome the deficiencies of the prior art and is a double-layer power grid comprehensive resource optimization configuration method considering power quality governance and load optimal scheduling. It can solve the steady-state power quality problems of low voltage, high harmonic content, and aggravated three-phase imbalance caused by the access of a large number of electric heating devices to the rural low-voltage distribution network. The method is simple and easy to implement.

[0005] Technical solution of the present invention: A comprehensive resource optimization allocation method after large-scale application of decentralized electric heating, characterized in that it comprises the following steps:

[0006] (1) Establish a two-layer power grid comprehensive resource optimization allocation model, including two parts: an upper-layer power quality governance device optimization allocation model and a lower-layer load optimization scheduling model;

[0007] (1-1) The upper-layer model is the power quality governance device optimization allocation model, with the minimum sum of the investment cost, operation and maintenance cost, transformer loss cost, and line loss cost of the power quality governance device as the objective function, and the decision variables being the location, quantity, and capacity of the installed power quality governance device. The objective function formula is shown in Equation (1).

[0008] T = min(ω1f V + ω2f cost + ω3f loss ) (1)

[0009] In the formula: f V , f cost , f loss are the transformer loss cost, the total cost of the power quality governance device, and the line loss cost respectively; ω1, ω2, and ω3 are the weight coefficients of the transformer loss cost, the total cost of the power quality governance device, and the line loss cost respectively.

[0010] (1-2) Calculate the transformer loss cost, the total cost of the power quality governance device, and the line loss cost in Equation (1) through Equations (2)-(4) respectively.

[0011]

[0012] f cost = C cost + C service = (C APF,cost + C CAP,cost + C SW,cost + C service ) (3)

[0013]

[0014] In the formula: C cost and C service are the total investment cost and the total operation and maintenance cost of the power quality governance device respectively;.C APF,cost , C CAP,cost , C SW,costThey are the investment costs of the thyristor switched capacitor (TSC), active power filter (APF), and low-voltage load on-line automatic phase conversion device for the governance device respectively; N is the number of line nodes; R i The equivalent resistance of branch i; P i , Q i , P loss They are the active power, reactive power of branch i, and the network loss of the distribution network system respectively; U i , U ref , ΔU are the voltage of node i, the reference voltage amplitude during normal operation, and the total voltage deviation respectively;

[0015] (1 - 3) The lower-layer model is the load optimal scheduling model. By adjusting the state of the three-phase unbalance adjustment device, the phases of the loads on each branch are adjusted, thereby reducing the three-phase unbalance of the distribution network; the lower-layer model takes the minimum three-phase unbalance of the line and the minimum adjustment times of the three-phase unbalance adjustment device as the objective function, and the objective function is shown in formula (5):

[0016] f = λ1I e +λ2m (5)

[0017] In the formula: λ1 is the weight coefficient of the three-phase unbalance; λ2 is the weight coefficient of the phase conversion switch number; I e is the three-phase unbalance of the distribution network; m is the adjustment times of the three-phase unbalance adjustment device;

[0018] (2) The particle swarm algorithm is used to obtain the optimal result of the upper-layer power quality governance device optimal configuration model;

[0019] The particle swarm algorithm solution process in step (2) specifically includes the following steps:

[0020] (2 - 1) Initialize the installation positions of the three-phase unbalance adjustment device and the active power filter (APF), and generate the initial position and velocity of the particle swarm; the coding method of the particle swarm is shown in formula (6). The first part of the particle represents whether the three-phase unbalance adjustment device is installed, represented by P i1 ~P iN . When the coding is 0, the three-phase unbalance adjustment device is not installed here. When the coding is 1, the three-phase unbalance adjustment device is installed; the second part of the particle represents the capacity of the TSC, represented by P iN+1 ; the third part of the particle represents the installed capacity of the APF, represented by P iN+2 ;

[0021] P i,PSO =[Pi1 …P iN ,P iN+1 ,P iN+2 ] (6)

[0022] (2-2) Calculate the fitness of each solution in the current population, that is, the objective function value;

[0023] (2-3) Update local optimal solution and global optimal solution;

[0024] (2-4) Update the position and velocity of each particle;

[0025] (2-5) Determine whether the iteration termination condition is met. If so, output the optimal solution. If not, return to (2-2).

[0026] (3) The optimized power quality control device location and capacity outputted from step (2) are fed into the lower-level load optimization scheduling model;

[0027] (4) Using genetic algorithm to obtain the optimal result of the lower-level load optimization scheduling model;

[0028] In step (4), the genetic algorithm solution process specifically includes the following steps:

[0029] (4-1) Chromosome encoding of the switch state of the three-phase imbalance adjustment device is performed according to the vector gene encoding strategy, and an initial population is generated;

[0030] The vector gene coding strategy in step (4-1) refers to a method in which the switching conditions of the phase-changing switches installed on the load branch correspond to a 3-bit binary code or a 3-dimensional column vector; if the phase-changing switch switches the load to phase A and does not switch to phase B or C, the corresponding bit of A can be represented as 1, and the corresponding bits of B and C can be represented as 0; the same applies if the phase-changing switch switches the load to phases B and C; then the switching conditions of each load branch can be represented by a switch phase sequence state column vector k as shown in formula (7):

[0031]

[0032] Therefore, the switching conditions of the phase-changing switches installed on all load branches can be represented by a switch phase sequence state matrix;

[0033] The phase sequence state matrices K0 and K in the low-voltage load online automatic phase-changing device before and after the switch phase sequence changes are:

[0034]

[0035] K=[k1,k2,k3...,k n ] (9)

[0036] Wherein: k i are the phase sequence state vectors before and after commutation of the i-th branch respectively;

[0037] (4-2) Calculate the fitness value of each individual according to the fitness function, that is, the objective function value;

[0038] (4-3) Perform genetic operations, select, cross, and mutate to generate an offspring population;

[0039] (4-4) Determine whether the iteration termination condition is satisfied. If satisfied, output the optimal solution. If not, return to (4-2);

[0040] (5) Apply the optimal positions and capacities of the power quality improvement devices and the optimal commutation control commands of the three-phase unbalance adjustment devices output in steps (2) and (4) to the selected distribution network to obtain the optimal configuration plan of the comprehensive resources of the regional distribution network.

[0041] The working principle of the present invention: The upper layer planning is the planning problem of the power quality improvement device. The planning goal is to minimize the investment cost, operation and maintenance cost, and transformer and line loss cost of the improvement device. The decision variables are the capacity configuration of the harmonic improvement device and the reactive power compensation device, and the number and position of the commutation switch configuration; the lower layer is the load optimal scheduling, with the goal of minimizing the number of operating switches and the three-phase unbalance degree. The decision variable is the state of the commutation switch. The upper layer planning result and the configuration of the power quality improvement device provide the initial conditions for the lower layer planning. The optimal operation value obtained from the lower layer planning is fed back to the upper layer planning, and the total cost value is recalculated to obtain the total objective function value of the upper layer planning.

[0042] The superiority of the present invention: A two-layer optimization model is adopted. The upper layer planning aims to minimize the investment cost, operation and maintenance cost, and distribution network loss cost of the improvement device; the lower layer is the load optimal scheduling, with the goal of minimizing the number of operating switches and the three-phase unbalance degree. After optimizing the grid structure, a two-layer planning model considering the configuration of the power quality improvement device and the load optimal adjustment of the three-phase automatic commutation device is established. The model not only considers the practicability and economy of installing the power quality improvement device, but also considers the load optimal adjustment to manage the three-phase unbalance problem of the substation area. A joint simulation platform based on Matlab and OpenDSS is built, and an improved particle swarm algorithm and genetic algorithm are used for multi-objective optimization solution, which is more convenient and faster in processing than the traditional simulation platform.

Description of the Drawings

[0043] Figure 1 It is a schematic flow chart of the comprehensive resource optimization configuration method after the large-scale application of a decentralized electric heating system involved in the present invention.

[0044] Figure 2 This is a flow chart for solving a particle swarm optimization algorithm involved in the present invention.

[0045] Figure 3 This is a flow chart for solving a genetic algorithm involved in the present invention.

[0046] Figure 4 This is a schematic diagram of the system structure built by the MATLAB and OpenDSS co-simulation platform used in the comprehensive resource optimization configuration method after the large-scale application of a decentralized electric heating system involved in the present invention.

[0047] Figure 5 This is a topology diagram of the low-voltage substation area in the comprehensive resource optimization configuration method after the large-scale application of a decentralized electric heating system involved in the present invention.

[0048] Figure 6 This is a diagram showing the voltage distribution at each node before and after the comprehensive resource optimization configuration under a 20 - 80% penetration rate in an embodiment involved in the present invention.

[0049] Figure 7 This is a harmonic spectrum analysis diagram at the point of common coupling before and after the comprehensive resource optimization configuration under a 20 - 80% penetration rate in an embodiment involved in the present invention.

Detailed implementation manners

[0050] In order to enable those skilled in the art to better understand the solution of this application, the following will further elaborate on this application in conjunction with the accompanying drawings.

[0051] Technical solution of the present invention: To solve the above technical problems, an embodiment of the present invention provides a comprehensive resource optimization configuration method after the large-scale application of a decentralized electric heating system. Figure 1 The following shows a schematic flow chart of a comprehensive resource optimization configuration method after the large-scale application of a decentralized electric heating system provided by an embodiment of the present invention. This method includes the following steps:

[0052] (1) Establish a double-layer power grid comprehensive resource optimization configuration model;

[0053] (2) Use the particle swarm optimization algorithm to obtain the optimal result of the upper-layer planning;

[0054] (3) Output the position and capacity of the optimized power quality improvement device to the lower-layer planning model;

[0055] (4) Use the genetic algorithm to obtain the optimal result of the lower-layer planning;

[0056] (5) Output the position and capacity of the optimal power quality improvement device of the double-layer model, and the optimal commutation control instruction of the three-phase unbalance adjustment device;

[0057] Furthermore, in the step (1), a two - layer integrated resource optimization configuration model for the power grid is established, where the upper - layer model is the optimization configuration model of power quality control devices. The upper - layer model takes the minimum of the investment cost, operation and maintenance cost of power quality control devices, and the transformer and line loss cost as the objective function. The decision variables are the location, quantity, and capacity of the installed power quality control devices. The objective function is shown in formulas (1)-(4):

[0058] T = min(ω1f V + ω2f cost + ω3f loss ) (1)

[0059]

[0060] f cost = minC cost + C service = min(C APF,cost + C CAP,cost + C SW,cost + C service ) (3)

[0061]

[0062] In the formula: C cost and C service are the total investment cost and the total operation and maintenance cost of the power quality control devices respectively;.C APF,cost , C CAP,cost , C SW,cost are the investment costs of the control devices which are Thyristor Switched Capacitor (TSC), Active power filter (APF), and the on - line automatic phase - changing device for low - voltage loads respectively; N is the number of line nodes; R i is the equivalent resistance of branch i; P i , Q i , P loss are the active power, reactive power of branch i and the network loss of the distribution network system respectively; U i , U ref , ΔU are the voltage of node i, the reference voltage amplitude under normal operation, and the total voltage deviation respectively;

[0063] Furthermore, in the step (1), the investment cost of the power quality control device includes three parts: the cost of APF, the cost of TSC, and the cost of the three - phase unbalance adjustment device. The investment cost of the power quality control device is obtained through formulas (5 - 7)

[0064]

[0065] C CAP,cost = minC cap Q C (6)

[0066] C SW,cost = minC SW N (7)

[0067] Where: S is the rated capacity of the active power filter, I Ahi and U are the harmonic current and voltage values respectively, b0 and b1 are the base price of 400,000 yuan for the active power filter and the coefficient of 450,000 yuan / MVA between the capacity and the cost; C cap is the investment cost per unit capacity, Q C is the installed capacity of the reactive power compensation device; C SW is the investment cost of a single phase change switch, and N is the number of phase change switches installed;

[0068] Furthermore, in the step (1), the operation and maintenance cost calculation formula of the power quality governance device is as follows:

[0069]

[0070] Where: C mat is the annual operation and maintenance cost per unit capacity of the device, λ mat is the ratio between the annual operation and maintenance cost and the initial investment cost; γ is the inflation rate, R is the social discount rate; T is the operation life of the device;

[0071] including equality constraints and inequality constraints:

[0072] Furthermore, in the step (1), the upper layer model objective function includes equality constraints and inequality constraints. The equality constraints are shown in equations (9) and (10).

[0073]

[0074]

[0075] Where: P Gi , P Li are the active power and active load of the generator at node i respectively; Q Gi , Q Li , Q Ci are the reactive power, reactive load and reactive compensation capacity of the generator at node i respectively; U i , U j are the voltages at node i and node j respectively; G ij , B ij , θ ijis the conductance, susceptance, and the phase angle difference of the node voltage between the connected nodes i and j; N is the total number of nodes in the system;

[0076] The inequality constraints are control variable constraints and state variable constraints. The generator node voltage and reactive power compensation capacity are selected as control variable constraints, and the node voltage, PCC power factor, total voltage harmonic distortion rate, harmonic current, and three-phase unbalance degree are selected as state variable constraints, as shown in formula (11):

[0077]

[0078] In the formula: U Gi , U Gimin , U Gimax are the break point voltage of generator node i and its lower and upper limit values respectively; Q C,H , Q C,H,min , Q C,H,max are the reactive power compensation capacity of the substation and its lower and upper limit values respectively; U Li , U Li,max , U Li,min are the voltage of node i and its lower limit value and upper limit value respectively; PF i is the fundamental power factor of the grid connection point during the operation of the electric heating user; HRU h is the content rate of the hth harmonic voltage, U h is the effective value of the hth harmonic voltage, U1 is the effective value of the fundamental voltage, U THD is the total distortion rate of the harmonic voltage, I Bh and I h0 are the harmonic current injected into the common connection point and the allowable value of the harmonic current, and e is the three-phase unbalance degree of the common connection point of the distribution network.

[0079] Furthermore, in the step (1), the lower-layer model is load optimal scheduling. By adjusting the state of the three-phase unbalance adjustment device, the phase of each branch load is adjusted, thereby reducing the three-phase unbalance degree of the distribution network. The lower-layer model takes the minimum three-phase unbalance degree of the line and the minimum adjustment times of the three-phase unbalance adjustment device as the objective function, and the objective function is shown in formula (12):

[0080] f = λ1I e % + λ2m (12)

[0081] In the formula: λ1 is the weight coefficient of the three-phase unbalance degree, and λ2 is the weight coefficient of the phase change switch number.

[0082] Furthermore, in the step (2-1), the encoding of the particle is as shown in formula (13), the first part of the particle indicates whether the three-phase imbalance adjustment device is installed, represented by , when the encoding is 0, the three-phase imbalance adjustment device is not installed here, and when the encoding is 1, the three-phase imbalance adjustment device is installed; the second part of the particle indicates the capacity of the TSC, represented by ; the third part of the particle indicates the installed capacity of the APF, represented by .

[0083] P i,PSO =[P i1 …P iN ,P iN+1 ,P iN+2 ] (13)

[0084] Furthermore, in the step (2-2), the fitness value of each particle is calculated, and the value corresponding to the decision variable of each particle needs to be obtained. Therefore, the distribution network corresponding to each particle state needs to be calculated for the power flow. The data corresponding to each particle is obtained by performing power flow calculation on the distribution network based on the MATLAB and OpenDSS joint simulation platform. Figure 4 The entire computing platform consists of three main modules: 1) distribution network three-phase power flow calculation module, 2) distribution network harmonic calculation module, and 3) integrated resource optimization configuration module. OpenDSS is used to calculate the three-phase power flow and harmonics of the distribution network, while integrated resource optimization configuration is implemented on the MATLAB platform. A component object (OpenDSSEngine.DLL) enables data communication between the OpenDSS program and other computational analysis modules on the MATLAB platform.

[0085] Furthermore, in the step (2-3), the fitness of each particle is compared with the fitness of the particle at the best position experienced. If the fitness is better, it is updated to the current local optimal solution.

[0086] Furthermore, in the step (2-3), the fitness of each particle is compared with the fitness of the particle at the best position experienced globally. If the fitness is better, it is updated to the current global optimal solution.

[0087] Furthermore, in the step (3), outputting the optimized power quality management device location and capacity to the lower-level planning model means installing the optimal power quality management device that has been obtained in the distribution network of the lower-level planning model, including installing it at the optimal location and installing its optimal capacity.

[0088] Further, in the step (4-1), the vector gene coding strategy refers to coding the states of the phase-change switches installed on the load branches. The specific coding strategy is as follows: The switching conditions of the phase-change switches installed on the load branches correspond to 3-bit binary coding or three-dimensional column vectors. If the phase-change switch switches the load to phase A (meanwhile, not switched to phase B or C), the corresponding bit of phase A can be represented as 1, and the corresponding bits of phases B and C are represented as 0. The same applies to switching to phases B and C. Then, the switching condition of each load branch can be represented by a switch phase sequence state column vector k, as shown in Equation (14):

[0089]

[0090] Therefore, the switching conditions of the phase-change switches installed on all load branches can be represented by a switch phase sequence state matrix. The phase sequence state matrices K0 and K before and after the change of the switch phase sequence in the low-voltage load on-line automatic phase-change device are respectively:

[0091]

[0092] K = [k1, k2, k3..., k n (16)

[0093] In the formula: k i are the phase sequence state vectors before and after phase change of the i-th branch respectively.

[0094] Further, in the step (4-2), to calculate the fitness of each individual, it is necessary to obtain the fitness value of each individual through power flow calculation of the distribution network on the joint simulation platform based on MATLAB and OpenDSS;

[0095] Further, in the step (4-3), the selection operation refers to selecting the main body of the next generation according to the principle of fitness selection. Individuals with higher fitness have a larger number of offspring in the next generation, and individuals with lower fitness have a smaller number of offspring in the next generation;

[0096] Further, in the step (4-3), the crossover operation refers to randomly selecting the same positions of two selected individuals for breeding the next generation and exchanging them at the selected positions according to the crossover probability;

[0097] Further, in the step (4-3), the mutation operation is based on the principle of gene mutation in biological inheritance, and mutations are performed on certain bits of certain individuals with a mutation probability. During mutation, the corresponding bits of the string to be mutated are inverted, that is, 1 is changed to 0 and 0 is changed to 1;

[0098] Furthermore, in step (4-4), when the fitness value of the optimal individual reaches a given threshold, or the fitness of the optimal individual and the fitness of the group no longer increase, the algorithm iteration process converges, the algorithm ends, and the optimal solution is output. Otherwise, the new generation of groups obtained through selection, crossover, and mutation replaces the previous generation of groups, and the loop returns to (4-2) to continue.

[0099] Furthermore, in step (5), the optimal position and capacity of the power quality management device of the double-layer model and the optimal phase-changing control instructions of the three-phase imbalance adjustment device output by steps (2) and (4) are applied to the selected distribution network to obtain the optimal configuration plan for the comprehensive resources of the distribution network in the region.

[0100] Application examples of the present invention:

[0101] 1. Basic parameters of the embodiment

[0102] The low-voltage distribution network of Caosi Village, Songzhuang Town, Tongzhou District, Beijing is selected. The short-circuit capacity of this low-voltage substation is 10MVA and it is powered by a 315kVA S11 distribution transformer. The network distribution diagram and topology diagram are as follows: Figure 5 As shown. The distribution network's power supply area has a radius of 681 meters and consists of five branches: two long branches measuring 302 meters and 213 meters, and two short branches measuring 115 meters, 190 meters, and 77 meters. The conductor type between each node is LGJ. There are 140 single-phase households in the substation area, distributed roughly evenly across each phase. Every household in the area has installed an air source heat pump, increasing the average electric heating capacity per household from 1kW to 5kW.

[0103] 2. Example Results

[0104] First, referring to steps (1) to (3), the upper-layer resources of the distribution network are optimized for four application scenarios with a penetration rate of 20% to 80% in the region, with the optimization goal of minimizing the total investment cost, operation and maintenance cost, transformer loss cost, and line loss cost of the power quality management device. The optimal TSC, APF, and three-phase commutation switch configuration for the distribution network resource optimization in the region are obtained as shown in Tables 1 and 2.

[0105] Table 1 TSC and APF configuration

[0106]

[0107]

[0108] Table 2 Three-phase commutation switch configuration

[0109]

[0110] Then, referring to steps (3) to (4), optimize the lower-layer resources of the regional distribution network, and the state positions of the three-phase phase-changing switches of the regional distribution network are obtained as shown in Table 3.

[0111] Table 3 Phase-changing switch positions

[0112]

[0113] Finally, referring to step (5), apply the optimal positions and capacities of the power quality governance devices and the optimal phase-changing control commands of the three-phase unbalance adjustment devices output by the double-layer model to the selected distribution network. The three-phase unbalance degrees before and after the comprehensive resources of the regional distribution network are obtained as shown in Table 4, and the voltage distribution is as Figure 6 shown, and the harmonic distribution is as Figure 7 shown.

[0114] Table 4 Statistical table of power quality evaluation indicators after power quality governance

[0115]

[0116] It can be seen from Table 4 that the three-phase unbalance degree and power factor of the regional distribution network have been greatly improved after the comprehensive resource optimization. Taking the scenario with a penetration rate of 80% as an example, the three-phase unbalance degree has dropped from 26.66% to 1.22%, a decrease of 20 times; the power factor has increased from 0.8648 to 0.9697, an increase of 0.1049. It can be concluded that the method for optimizing the allocation of distribution network resources proposed by the present invention can effectively improve the three-phase unbalance degree and power factor of the distribution network.

[0117] From Figure 6 it can be seen that the voltages of each node in the regional distribution network have been greatly improved after the comprehensive resource optimization. Among them, the minimum per-unit values of the voltages in different application scenarios with penetration rates of 20% to 80% are 0.9745 p.u., 0.9317 p.u., 0.9017 p.u., and 0.9486 p.u., respectively, which have met the requirements of the low-voltage distribution network limit of +7% to -10%. It can be concluded that the method for optimizing the allocation of distribution network resources proposed by the present invention can effectively improve the voltage sag situation of the distribution network. Figure 6 Among them, they respectively represent the comparison effects of the node voltages before and after optimization when the penetration rates are 20%, 40%, 60%, and 80%.

[0118] From Figure 7It can be seen that after the comprehensive resource optimization of the regional distribution network, the harmonic pollution has been greatly improved. The improvement effect of the total voltage harmonic distortion rate at the PCC point is better. At the same time, the harmonic voltage content rates under different application scenarios with a penetration rate of 20% - 80% are reduced to 1.05%, 1.38%, 0.91% and 0.86% respectively, all meeting the national standard limit of the total voltage harmonic distortion rate < 4%. It can be concluded that the method for optimizing the allocation of distribution network resources proposed by the present invention can effectively improve the harmonic pollution problem of the distribution network.

[0119] In summary, the method for optimizing the allocation of comprehensive resources after the large-scale application of decentralized electric heating proposed by the present invention can effectively control the power quality problem of the distribution network after the large-scale application of electric heating, while taking into account the economy and the power quality control effect. The proposed double-layer resource optimization model comprehensively considers the practicability and economy of installing power quality control devices.

Claims

1. A comprehensive resource optimization allocation method after large-scale application of decentralized electric heating, characterized in that It includes the following steps: (1) Establish a comprehensive resource optimization configuration model for the double-layer power grid, including two parts: the upper-layer optimal configuration model of power quality governance devices and the lower-layer load optimal scheduling model; (1-1) The upper-layer model is the optimal configuration model of power quality governance devices. The objective function is to minimize the sum of the investment cost, operation and maintenance cost, transformer loss cost, and line loss cost of power quality governance devices. The decision variables are the location, quantity, and capacity of power quality governance devices; The formula of the objective function is shown in Equation (1): T = min(ω1f V + ω2f cost + ω3f loss ) (1) where: f V , f cost , f loss are the transformer loss cost, the total cost of the power quality control device, and the line loss cost respectively; ω1, ω2, and ω3 are the weight coefficients of the transformer loss cost, the total cost of the power quality control device, and the line loss cost respectively; (1-2) Calculate the transformer loss cost, total cost of power quality governance devices, and line loss cost in Equation (1) through Equations (2)-(4) respectively; f cost = C cost + C service = (C APF,cost + C CAP,cost + C SW,cost + C service ) (3) Where: C cost and C service are respectively the total investment cost and the total operation and maintenance cost of the power quality control device; .C APF,cost , C CAP,cost , C SW,cost are the investment costs of the thyristor switched capacitor, active power filter, and low-voltage load online automatic phase conversion device of the governance device respectively; N is the number of line nodes; R i is the equivalent resistance of branch i; P i , Q i , P loss are the active power, reactive power of branch i, and the network loss of the distribution network system respectively; U i , U ref , ΔU are the voltage of node i, the reference voltage amplitude during normal operation, and the total voltage deviation respectively; (1-3) The lower-layer model is the load optimal scheduling model. By adjusting the state of the three-phase unbalance regulating device, the phase of each branch load is adjusted, thereby reducing the three-phase unbalance of the distribution network; The objective function of the lower-layer model is to minimize the three-phase unbalance of the line and the number of regulation times of the three-phase unbalance regulating device. The objective function is shown in Equation (5): f = λ1I ε + λ2m (5) Where: λ1 is the weight coefficient of the three-phase unbalance degree; λ2 is the weight coefficient of the number of commutation switches; I ε is the three-phase unbalance degree of the distribution network; m is the number of adjustment times of the three-phase unbalance adjustment device; (2) Use the particle swarm optimization algorithm to obtain the optimal result of the upper-layer optimal configuration model of power quality governance devices; (3) Output the optimized location and capacity of power quality governance devices in step (2) to the lower-layer load optimal scheduling model; (4) Use the genetic algorithm to obtain the optimal result of the lower-layer load optimal scheduling model; (5) Apply the optimal location and capacity of power quality governance devices and the optimal phase change control instruction of the three-phase unbalance regulating device output in steps (2) and (4) to the selected distribution network to obtain the optimal configuration plan for the comprehensive resources of the regional distribution network.

2. The comprehensive resource optimization configuration method after large-scale application of a decentralized electric heating system according to claim 1, wherein The specific solution process of the particle swarm optimization algorithm in step (2) specifically includes the following steps: (2-1) Initialize the installation positions of the three-phase unbalance adjustment device and the active power filter, and generate the initial positions and velocities of the particle swarm; the coding method of the particle swarm is shown in Equation (6). The first part of the particle represents whether the three-phase unbalance adjustment device is installed, denoted by P i1 ~P iN . When the code is 0, the three-phase unbalance adjustment device is not installed here; when the code is 1, the three-phase unbalance adjustment device is installed. The second part of the particle represents the capacity of the TSC, denoted by P iN+1 . The third part of the particle represents the installed capacity of the APF, denoted by P iN+2 . P i,PSO = [P i1 …P iN , P iN+1 , P iN+2 (6) (2-2) Calculate the fitness of each solution in the current population, that is, the objective function value; (2-3) Update the local optimal solution and the global optimal solution; (2-4) Update the position and velocity of each particle; (2-5) Judge whether the iteration termination condition is satisfied. If satisfied, output the optimal solution. If not satisfied, return to (2-2).

3. The comprehensive resource optimization configuration method after large-scale application of a decentralized electric heating according to claim 1, characterized in that In step (4), the specific solution process of the genetic algorithm specifically includes the following steps: (4-1) Encode the switch state of the three-phase unbalance regulating device according to the vector gene encoding strategy and generate an initial population; (4-2) Calculate the fitness value of each individual according to the fitness function, that is, the objective function value; (4-3) Perform genetic operations, select, cross, and mutate to generate an offspring population; (4-4) Judge whether the iteration termination condition is satisfied. If satisfied, output the optimal solution. If not satisfied, return to (4-2).

4. The comprehensive resource optimization configuration method after large-scale application of a decentralized electric heating system according to claim 3, characterized in that The vector gene encoding strategy in the step (4-1) refers to the method in which the switching conditions of the phase-change switches installed on the load branches correspond to 3-bit binary codes or three-dimensional column vectors; if the phase-change switch switches the load to phase A, and at the same time, does not switch to phase B or C, the corresponding bit of phase A can be represented as 1, and the corresponding bits of phases B and C are represented as 0; the same applies when the phase-change switch switches the load to phase B or C; then the switching condition of each load branch can be represented by a switching phase sequence state column vector k as shown in formula (7): Therefore, the switching conditions of the phase-change switches installed on all load branches can be represented by a switching phase sequence state matrix; The phase sequence state matrices K0 and K before and after the phase sequence of the switching phase sequence in the low-voltage load on-line automatic phase-change device remains unchanged are respectively: K = [k1, k2, k3..., k n (9) Wherein: k i are the phase sequence state vectors before and after commutation of the i-th branch respectively.

Images