CN113673065B - Loss reduction method for automatic reconstruction of power distribution network - Google Patents
Loss reduction method for automatic reconstruction of power distribution network Download PDFInfo
- Publication number
- CN113673065B CN113673065B CN202110925676.1A CN202110925676A CN113673065B CN 113673065 B CN113673065 B CN 113673065B CN 202110925676 A CN202110925676 A CN 202110925676A CN 113673065 B CN113673065 B CN 113673065B
- Authority
- CN
- China
- Prior art keywords
- network
- loss
- node
- power distribution
- distribution network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/25—Design optimisation, verification or simulation using particle-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/04—Power grid distribution networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- Mathematical Analysis (AREA)
- Biophysics (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Pure & Applied Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a loss reduction method for automatic reconstruction of a distribution network, which overcomes the defects of insufficient automation of the distribution network and the problems of economic operation of the network under the coexistence of light load and heavy load in the prior art, and comprises the following steps: detecting network data of the power distribution network, and sorting and encoding the data; calculating the line loss on each branch in the power distribution network by using a forward-backward substitution method; carrying out 0/1 coding on each section of the current network, and then constructing a target function; automatically optimizing and adjusting the operation mode of the power distribution network by utilizing a particle swarm optimization algorithm; calculating the line loss and the distribution transformer loss after the network reconstruction by a forward-backward substitution method; and comparing the line loss and the distribution transformer loss before and after the reconstruction of the whole looped network switch. A network reconstruction decision-making model of a typical power supply unit of a power distribution network for low-loss economic operation is provided, the operation mode is dynamically adjusted through power distribution network automation, the problem of network economic operation under the condition of light load and heavy load is systematically solved, and the reduction of the network loss rate is realized.
Description
Technical Field
The invention relates to the technical field of optimization of a power distribution network of a power system, in particular to a loss reduction method for automatic reconstruction of a power distribution network.
Background
The line loss refers to the loss of electric energy generated in each link of power transmission, power transformation, power distribution, sale and the like in the process of transmitting the electric energy from a power plant to a power consumer. The line loss in the power grid accounts for a considerable part of the total power generation, so that the reduction of the line loss has important practical significance. The problems that at present, heavy load and light load coexist, the load is increased unevenly, and the seasonal nature of power utilization is obvious are caused to the operation of the power distribution network. In the long-term distribution network operation management process, more attention is paid to the heavy load problem, the safety problem under the heavy load operation is mainly solved by the power grid through the lifting equipment level, great effect is achieved, and the reliability of power supply of the distribution network is obviously improved. But also causes larger net rack redundancy, the utilization efficiency of the equipment is lower, and the light load problem is prominent. Therefore, in order to achieve the goal of further reducing the grid loss rate, the problems of heavy load and light load loss must be systematically concerned.
Distribution network automation is an effective auxiliary means for improving the operation safety of a distribution network, but most of the application scenes of the current distribution network automation of the power grid are concentrated under a low-probability fault state to supply fault area loads, and the utilization rate of the distribution automation is low due to the fact that faults have natural low-probability attributes. The essence of the distribution automation is that the global automatic switching of the switch equipment is realized, and the distribution automation has full-period flexible switching capability on the network frame topology in the operation process of the distribution network, and is suitable for fault states. The large-scale load cutting and line transformation are limited, so that the optimized operation based on the existing net rack becomes the main direction for improving the operation efficiency of the power distribution network and reducing the line loss rate.
The chinese patent office 2021, 5 months and 28 days discloses an invention named as a power distribution network energy-saving loss-reducing system and method based on big data, and the publication number is CN112862243A, and the invention system comprises: the power grid power consumption reduction system comprises a power grid voltage detection module, a power grid current detection module, a power grid power detection module, a central control module, a power consumption calculation module, a loss reduction rate calculation module, a power grid energy-saving optimization module, an energy-saving loss reduction evaluation module, a big data processing module and a display module. According to the invention, the workload of related personnel is greatly reduced through the loss reduction rate calculation module; meanwhile, the energy-saving loss-reducing evaluation module considers the correction of the non-parameter characteristics of the line on the safety loss, the hour is taken as a time window, the calculation precision of the period cost and the loss is improved, the line adaptability is evaluated according to the economic and energy-saving two-dimensional standard, and an energy-saving loss-reducing adaptability evaluation basis is provided for the existing line and line planning and type selection. But no mention is made of distribution network automation, nor is line loss reduced from the point of view of network reconfiguration.
Disclosure of Invention
The invention aims to provide a loss reduction method for automatic reconstruction of a power distribution network aiming at the defects of the existing scheme, provides a network reconstruction decision model of a typical power supply unit of the power distribution network for low-loss economic operation, systematically solves the problem of network economic operation under the condition of light load and heavy load by dynamically adjusting the operation mode of the power distribution network in an automatic manner, and realizes the reduction of the network loss rate.
In order to achieve the purpose, the invention adopts the following technical scheme: a loss reduction method for automatic reconstruction of a power distribution network is characterized by comprising the following steps:
s1: detecting network data of the power distribution network, and sorting and encoding the data;
s2: calculating the line loss on each branch in the power distribution network by using a forward-backward substitution method;
s3: carrying out 0/1 coding on each section of line of the current network, wherein 0 represents that the line is disconnected, and 1 represents that the line is closed, and then constructing a target function;
s4: automatically optimizing and adjusting the operation mode of the power distribution network by utilizing a particle swarm optimization algorithm;
s5: calculating the line loss and the distribution transformation loss after the network reconstruction by a forward-backward substitution method;
s6: and comparing the line loss and the distribution transformer loss before and after the reconstruction of the whole looped network switch, and judging whether the adjusted network structure reduces the line loss or not.
The line loss rate and the load rate of the power distribution equipment represented by the distribution transformer have a dynamic mapping relation, and the line loss rate is increased due to overhigh and overlow load rates, so that the line loss rate and the distribution transformer load need to be optimally adjusted and actively managed, the load rate of the equipment is in an economic operation interval, and the purpose of overall loss reduction is achieved. On the other hand, in the conventional power distribution network operation management, the division of heavy load/light load is relatively extensive, generally, based on typical daily load actual measurement, lines and distribution transformers with load rates above a certain threshold value are simply divided into heavy loads, and the lines and distribution transformers below the certain threshold value are classified as light loads. The method does not consider the time attribute of equipment overload, and actual power equipment has short-time overload capacity, so that load historical data mining and equipment overload capacity analysis need to be carried out, a load rate-duration relation curve of typical equipment is drawn, and the cumulative time reaching a certain load rate and being above the load level reaches xx hours/year serving as a scientific criterion for equipment overload. Therefore, the utilization rate of equipment is scientifically evaluated, and deep sleep resources are accurately identified. The invention provides a low-loss economic operation-oriented network reconstruction decision-making model of a typical power supply unit of a power distribution network, which utilizes a particle swarm optimization algorithm, searches an optimal switch switching mode by taking line loss (line loss and distribution transformation loss) as a fitness function of the particle swarm optimization algorithm, analyzes and compares the line loss and the distribution transformation loss before and after the switch reconstruction of the whole ring network, thereby finding the minimum line loss operation mode of the ring network, systematically solving the network economic operation problem under the coexistence of light load and heavy load, realizing the reduction of the network loss rate on the one hand, releasing the automation of the power distribution network and the flexibility potential of the existing network frame on the other hand, and realizing the purposes of awakening sleeping resources and enabling asset increment on the other hand.
Preferably, the network data collected in step S1 includes: the number of lines, the number of nodes, the number of switching stations, the number of substations and the voltage, current and power are coded by digital coding. Encoding is performed to facilitate post-processing and labeling.
Preferably, in step S2, the calculation method for calculating the line loss on each branch in the power distribution network by using the push-back method is as follows:
s2.1: inputting network data, searching network nodes and forming a hierarchical relationship;
s2.2: initializing the node voltage:
U i ＝1(i＝0,1,2,3,…,n)
in the formula of U 0 The voltage of the node is 0 point voltage, namely the voltage of the root node of the tree, and n is the number of the nodes;
s2.3: carrying out back substitution calculation: calculating the current of each branch upwards layer by layer from the last 1 layer of load nodes through the layering of the tree-shaped network;
s2.4: and (3) performing forward calculation: calculating from the root node to the 1 st layer from the 1 st layer to the 1 st layer, and calculating the voltage value of each node layer by layer;
s2.6: according to the formula:
determine whether to converge, wherein U i (t) node Voltage, U, derived from the present Forward calculation i (t-1) nodes obtained by last forward calculationVoltage, ζ represents the calculation accuracy, if convergence, the line loss on each branch is calculated, and if not, steps S2.3-S2.5 are repeated until convergence.
The forward-backward substitution method is the prior art, meets the requirements of an actual power distribution network structure, does not need to calculate a node admittance matrix, is high in calculation efficiency, and can realize the radial power distribution network load flow calculation through multiple iterations.
Preferably, in step S3, the constructed objective function is the minimum overall loss, including the line loss and the distribution transformation loss, and the objective function may be represented as:
wherein, beta is a load factor,
is the load power factor, S N Rated capacity of the transformer, P 0 Denotes the power of the root node of the tree, k denotes the child node, P k Denotes the power of the child node, l denotes the number of branches, R l Denotes the resistance of the l-th branch, P' l1,2 Representing active power, Q ', between node 1 and node 2 on the branch' l1,2 Represents the reactive power, U ', between node 1 and node 2 on the branch' l1,2 Representing the voltage between node 1 and node 2 on the branch.
The target function of network reconstruction is the minimum overall loss, including line loss and distribution transformer loss, the calculation range is from the substation outlet of the looped network power supply unit to the low-voltage side bus of the distribution transformer, and the calculation covers the feeder loss from the substation outlet to the looped network cabinet/switching station, the loss from the looped network cabinet/switching station to the distribution transformer line and the loss from the distribution transformer. The invention takes a line switch and a distribution transformer low-voltage side bus-coupled switch as switchable equipment, the switch state of the switching equipment is expressed as a variable of 0-1, and the sum of line loss and distribution transformer loss is taken as a target function, so that an integer nonlinear programming model for the most economic operation of a looped network unit is established. And solving the model by combining a particle swarm optimization algorithm with a forward-push-back tidal current flow calculation.
Preferably, in the step S4, the specific step of automatically optimizing and adjusting the operation mode of the power distribution network by using the particle swarm optimization algorithm is as follows:
s4.1: setting initial parameters;
s4.2: forming a topological structure, and calculating the line loss and the total distribution transformer loss by a forward-backward substitution method;
s4.3: evaluating and sequencing the fitness of the particles;
s4.4: updating global extrema according to particle fitness
And individual extremum
And then updating all particle states, wherein the global extreme value is as follows:
where t represents the number of iterations, m represents the dimension,
representing the optimal position found by the whole particle swarm in the t-th iteration;
the individual extrema are:
in the formula (I), the compound is shown in the specification,
representing the optimal position found by the t iteration particle i;
the updated particle positions are:
wherein K is a convergence factor, r 1 And r 2 Is distributed in [0,1 ]]A random number in between; c. C 1 And c 2 In order to learn the factors, the learning device is provided with a plurality of learning units,
representing the velocity of the ith particle at the time of the t-th iteration,
represents the velocity of the ith particle at the t +1 th iteration, J being a constant;
when the iteration is carried out for t times, the j-th optimal solution searched by the particle swarm is obtained;
s4.5: and judging whether the fitness function termination condition is met, if so, finishing network optimization adjustment, and if not, repeating the steps S4.2-S4.5 until the fitness function termination condition is met.
The particle swarm optimization algorithm is a prior art, is proposed in 1995, is an evolutionary computing technology based on a swarm intelligence method, and is a method for searching an optimal solution through iteration starting from a group of random solutions. In particle swarm optimization, each "particle" represents a potential solution to the optimization problem, and all particles contain a fitness that is determined by the optimization function and a velocity that determines their flight direction and distance, and update their own velocity by following the currently sought optimal particle. The velocity of each particle consists of three components: the first part is the original speed of the particles and is the embodiment of the memory function of the particles; the second part is the relationship between the particles and the individual extremum; the third part of particles is related to the global extremum. In the invention, the states of all switchable switches are used as decision variables, 0-1 coding is adopted, forward-backward substitution load flow calculation is carried out on the network state corresponding to each particle, and the corresponding network loss value is used as example fitness.
Preferably, in step S4.1, the set initial parameters include: set algebra t =0, initial scale N 0 Setting control parameters and network topology dataAnd forming an initial population and an initial particle velocity.
Preferably, in step S4.3, the particle fitness is: and (3) carrying out forward-pushing back substitution load flow calculation on the network state corresponding to each particle, and taking the corresponding network loss value as an example fitness:
the line loss (line + distribution transformer) is used as a fitness function of the particle swarm algorithm, an optimal switch switching mode is found, and the line loss and the distribution transformer loss before and after the switch reconfiguration of the whole ring network are analyzed and compared, so that the minimum line loss operation mode of the ring network can be found.
Preferably, in step S6, comparing the line loss and the distribution transformer loss before and after the reconfiguration of the entire ring network switch, and determining whether the adjusted network structure reduces the line loss specifically is represented as:
and (3) comparing the line loss value obtained in the step (S2) with the line loss value obtained in the step (S5), judging whether the optimized and adjusted network topology structure reduces the line loss, if so, recording the network load state and the current network optimization mode, and if not, returning to the step (S3) to optimize and adjust the network again.
And analyzing and judging whether the line loss is reduced or not after the network is reconstructed so as to judge whether the method is practical or not.
Therefore, the invention has the following beneficial effects: 1. the invention reduces the line loss rate, thereby improving the operation efficiency of the power grid and reducing the power consumption cost of users; 2. the invention provides a network reconstruction decision-making model of a typical power supply unit of a power distribution network for low-loss economic operation, and systematically solves the problem of network economic operation under the coexistence of light load and heavy load by dynamically adjusting the operation mode of the power distribution network automation, so that the reduction of the network loss rate is realized; 3. the automation of the power distribution network and the flexibility of the existing network frame are improved, and the full utilization of resources is realized; 4. the load balance of the low-voltage distribution network can be automatically implemented to reduce line loss.
Drawings
FIG. 1 is a flow chart of the operation of the method of the present invention;
FIG. 2 is a numbering diagram of a heavy-duty distribution transformer load node;
FIG. 3 is a numbering diagram of a heavy-duty distribution transformer load node after network reconfiguration;
FIG. 4 is a line flow before and after reconfiguration of a heavy-duty distribution and transformation load distribution network;
FIG. 5 is a node numbering diagram of a light-load distribution transformer;
fig. 6 is a light-load distribution transformer load node numbering diagram after network reconfiguration;
FIG. 7 is a line flow before and after reconfiguration of a light-load distribution and transformation load distribution network;
in the figure: 001. a power source; 002. a switching station.
Detailed Description
The invention is described in further detail below with reference to the following figures and embodiments:
in the embodiment shown in fig. 1, a loss reduction method for automatically reconstructing a power distribution network can be seen, and the operation flow is as follows:
the first step is as follows: detecting network data of power distribution network, sorting and coding data
The collected network data includes: the number of lines, the number of nodes, the number of switchgears, the number of substations, and the voltage, current, power. Nodes and lines are encoded using digital coding to facilitate post-processing and labeling.
The second step is that: calculating line loss on each branch in a power distribution network using a push-forward back substitution method
The forward-backward substitution method is the prior art, meets the requirements of an actual power distribution network structure, does not need to calculate a node admittance matrix, is high in calculation efficiency, and can realize the radial power distribution network load flow calculation through multiple iterations.
The method comprises the following specific steps:
inputting network data, searching network nodes and forming a hierarchical relationship;
initializing the node voltage:
U i ＝1(i＝0,1,2,3,…,n)
in the formula of U 0 The voltage of the node is 0 point voltage, namely the voltage of the root node of the tree, and n is the number of the nodes;
carrying out back substitution calculation: calculating the current of each branch upwards layer by layer from the last 1 layer of load nodes through the layering of the tree-shaped network; and then forward calculation is carried out: starting from the root node, calculating from the layer 1 to the layer 1, and calculating the voltage value of each node layer by layer downwards; according to the formula:
judging whether convergence is present, wherein U i (t) node Voltage, U, derived from the present Forward calculation i And (t-1) calculating the node voltage obtained by the previous forward calculation, wherein zeta represents the calculation precision, if convergence occurs, the line loss on each branch is calculated, if convergence does not occur, the back substitution calculation and the forward calculation are carried out again, and whether convergence occurs or not is judged again until convergence occurs.
The third step: carrying out 0/1 coding on each section of line of the current network, wherein 0 represents that the line is open, 1 represents that the line is closed, and then constructing an objective function
The target function of network reconstruction is the minimum overall loss, including line loss and distribution transformer loss, the calculation range is from the substation outlet of the looped network power supply unit to the low-voltage side bus of the distribution transformer, and the calculation covers the feeder loss from the substation outlet to the looped network cabinet/switching station, the loss from the looped network cabinet/switching station to the distribution transformer line and the loss from the distribution transformer. The objective function of the optimization problem is the power supply unit voltage level loss, which can be expressed as:
wherein, beta is a load factor,
as the load power factor, S N Rated capacity of the transformer, P 0 Representing the power of the root node of the tree, k representing the childNode, P k Denotes the power of the child node, l denotes the number of branches, R l Denotes the resistance of the first branch, P' l1,2 Representing active power, Q ', between node 1 and node 2 on the branch' l1,2 Representing reactive power, U ', between node 1 and node 2 on the branch' l1,2 Representing the voltage between node 1 and node 2 on the branch.
The fourth step: automatic optimization adjustment of power distribution network operation mode by utilizing particle swarm optimization algorithm
The particle swarm optimization algorithm is a prior art, is proposed in 1995, is an evolutionary computing technology based on a swarm intelligence method, and is a method for searching an optimal solution through iteration starting from a group of random solutions. In particle swarm optimization, each "particle" represents a potential solution to the optimization problem, and all particles contain a fitness that is determined by the optimization function and a velocity that determines their flight direction and distance, and update their own velocity by following the currently sought optimal particle. The velocity of each particle consists of three components: the first part is the original speed of the particles and is the embodiment of the memory function of the particles; the second part is the relationship between the particles and the individual extremum; the third part of particles is related to the global extremum. In the invention, the states of all switchable switches are used as decision variables, 0-1 coding is adopted, forward-backward substitution load flow calculation is carried out on the network state corresponding to each particle, and the corresponding network loss value is used as example fitness.
Set algebra t =0, initial scale N 0 Setting control parameters and network topology data to form an initial population and a particle initial speed; forming a topological structure, and calculating the line loss and the total distribution transformer loss by a forward-backward substitution method; and evaluating and sequencing the fitness of the particles, wherein the fitness of the particles is as follows: forward-pushing and backward-replacing the load flow calculation on the network state corresponding to each particle, and taking the corresponding network loss value as an example fitness:
then, the global is updated according to the particle fitnessExtreme value
And individual extremum
And then updating all particle states, wherein the global extreme value is as follows:
where t represents the number of iterations, m represents the dimension,
representing the optimal position found by the whole particle swarm in the t-th iteration;
the individual extrema are:
in the formula (I), the compound is shown in the specification,
representing the optimal position found by the t iteration particle i;
the updated particle positions are:
wherein K is a convergence factor, r 1 And r 2 Is distributed in [0,1 ]]A random number in between; c. C 1 And c 2 In order to learn the factors, the learning device is provided with a plurality of learning units,
representing the velocity of the ith particle at the time of the t-th iteration,
indicates the ith granuleThe speed of the child at the t +1 th iteration, J being a constant;
when iteration is carried out for t times, j bit optimal solution searched by the particle swarm is obtained; and finally, judging whether the fitness function termination condition is met, if so, finishing network optimization adjustment, if not, reforming a topological structure, calculating line loss, evaluating and sequencing the fitness of the particles, calculating the positions of the updated particles, and judging whether the fitness function termination condition is met until the termination condition is met. The termination condition is set according to actual conditions.
The fifth step: calculating line loss and distribution transformer loss after network reconstruction by forward-backward substitution method
And a sixth step: comparing the line loss and the distribution transformer loss before and after the reconstruction of the whole looped network switch, and judging whether the adjusted network structure reduces the line loss
Comparing the line loss and the distribution transformer loss before and after the reconstruction of the whole looped network switch, and judging whether the adjusted network structure reduces the line loss specifically as follows:
and comparing the line loss value obtained in the second step with the line loss value obtained in the fifth step, judging whether the optimized and adjusted network topology structure reduces the line loss, if so, recording the network load state and the current network optimization mode, and if not, returning to the third step, and optimizing and adjusting the network again.
In the embodiment shown in fig. 2-4, a heavy-duty distribution transformer load is selected for analysis:
data sorting is carried out on the power distribution network, and digital numbers are adopted to encode nodes and lines so as to facilitate post-processing and marking. The specific node numbering diagram is shown in fig. 2, wherein the node numbers of the switchyard hitching and distribution transformer are not marked. According to the equivalent topological graph and the network data, the transformer substation and the switch station are used as main nodes, and information of each node and each line is obtained:
| number of | Head end node | End | Branch impedance | |
| 1 | 1 | 2 | 0.077059 | |
| 2 | 2 | 3 | 0.050664 | |
| 3 | 3 | 4 | 0.045436 | |
| 4 | 2 | 5 | 0.032648 | |
| 5 | 2 | 15 | 0.013624 | |
| 6 | 2 | 16 | 0.150388 | |
| 7 | 2 | 17 | 0.150388 | |
| 8 | 5 | 18 | 0.059736 | |
| 9 | 5 | 19 | 0.12844 | |
| 10 | 5 | 20 | 0.06856 | |
| 11 | 5 | 21 | 0.04325 | |
| 12 | 3 | 22 | 0.046112 | |
| 13 | 3 | 23 | 0.06026 | |
| 14 | 3 | 24 | 0.11528 | |
| 15 | 3 | 25 | 0.262524 | |
| 16 | 3 | 26 | 0.158248 | |
| 17 | 4 | 27 | 0.11956 | |
| … | … | … | … |
And performing forward-backward flow-replacing calculation on the circuit to obtain the line loss on each branch. Then each section of the line is subjected to 0/1 coding, wherein 0 represents that the line is open, and 1 represents that the line is closed. And the line loss minimum is defined as a fitness function by the particle swarm algorithm, and the operation mode of the particle swarm algorithm is optimized and adjusted to reconstruct the network.
The reconstructed node map is shown in fig. 3, before reconstruction, the loss of the whole looped network unit (line + distribution transformer) is 83.74kW (where the line loss is 23.53kW, and the distribution transformer loss is 60.21 kW), and after reconstruction, the looped network loss is recalculated by a forward-backward substitution method, and the line loss is reduced to 75.47kW (where the line loss is 15.26kW, and the distribution transformer loss is 60.21 kW). The operation mode optimization front and back line power flow distribution pairs are shown in figure 4. The loss of the looped network unit before and after the adjustment of the operation mode is shown as the following table:
| looped network unit data statistics | Before load adjustment | After the load is adjusted |
| Active load of looped network | 4290.5kW | 4290.5kW |
| Line loss of looped network line | 23.53kW | 15.26kW |
| Line loss of ring network transformer | 60.21kW | 60.21kW |
| Line loss of looped network | 83.74kW | 75.47kW |
| Line loss rate of looped network line | 0.55％ | 0.36％ |
| Line loss rate of ring network transformer | 1.40％ | 1.40％ |
| Loop network line loss rate | 1.95％ | 1.76％ |
By changing the operation mode, the line loss rate of the maximum load daily ring network is reduced to 1.76 percentage points from the original 1.95 percentage points, and is reduced by 0.19 percentage points, wherein the line loss rate is reduced by 0.11 percentage points. Therefore, the operation parameters are adjusted by changing the operation mode of the system, so that the power distribution of the network is close to economic distribution, the network operation is more economic, and the line loss is less, so that the looped network operation mode based on the distribution automation equipment has a better optimization effect on the overload line loss.
Now, assuming that each distribution transformer load is an annual average load and the running time is half of the annual total time (8760 h), the annual power consumption of the double-loop network of the group of cables is about 36.65 kWh, and the power consumption can be reduced by 3.57 kWh through the reconfiguration of the distribution network.
In the embodiment shown in fig. 5-7, a light load distribution transformer load is selected for analysis:
data arrangement is carried out on the power distribution network, and digital numbers are adopted to code nodes and lines so as to facilitate post-processing and marking. The specific node numbering diagram is shown in fig. 5, wherein the node numbers of the switchyard hitching and distribution transformer are not marked.
According to the topological graph and the network data, the transformer substation and the switch station are used as main nodes, and information of each node and each line can be obtained:
and (3) coding each section of line by 0/1 according to the existing network, wherein 0 represents that the line is open, and 1 represents that the line is closed. And the line loss minimum is defined as a fitness function by the particle swarm algorithm, and the operation mode of the ring network unit is optimized and adjusted by the particle swarm algorithm.
The reconstructed node map is shown in fig. 6. Before reconstruction, the loss of the whole looped network unit (line + distribution transformer) is 27.64kW (wherein the line loss is 0.8kW, and the distribution transformer loss is 26.84 kW), and after reconstruction, the looped network loss is recalculated through a push-back substitution method, and the line loss is reduced to 26.19kW (wherein the line loss is 0.7kW, and the distribution transformer loss is 26.19 kW). The operation mode optimization front and back line power flow distribution pairs are shown in fig. 7. The loss of the looped network unit before and after the operation mode is adjusted is shown as the following table:
| looped network unit data statistics | Before load adjustment | After the load is adjusted |
| Active load of looped network | 1110.235kW | 1110.235kW |
| Line loss of looped network line | 0.8kW | 0.7kW |
| Line loss of ring network transformer | 26.84kW | 25.49kW |
| Line loss of looped network | 27.64kW | 26.19kW |
| Line loss rate of looped network line | 0.08％ | 0.06％ |
| Line loss rate of ring network transformer | 2.41％ | 2.30％ |
| Ring network line loss rate | 2.49％ | 2.36％ |
In the light load state, the operation mode is not changed, only the position of the hanging distribution transformer needs to be adjusted, the line loss rate of the maximum load daily ring network is reduced to 2.36 percentage points from the original 2.49 percentage points, the line loss rate is reduced by 0.13 percentage point, and the loss rate of the distribution transformer is reduced by 0.11 percentage point. Therefore, the non-economic operation of the transformer can be effectively reduced by adjusting the switching mode of the transformer, the transformer loss is reduced, and the line loss is reduced, so that the transformer is adjusted based on the distribution automation equipment, and the light-load line loss is better optimized.
Now, assuming that each distribution transformer load is an average load all the year around, and the running time is half of the total time (8760 h) all the year around, the annual line power loss of the double-ring network of the group of cables is about 12.1 ten thousand kWh, and the power loss can be reduced by 0.63 ten thousand kWh through the reconfiguration of the distribution network.
Therefore, by means of operation mode adjustment, the problem of network economic operation under the condition of both light load and heavy load can be solved systematically, and meanwhile, the utilization rate of distribution automation is improved. The aim of reducing the line loss can be achieved only by adjusting the operation mode of the transformer without changing the operation mode of the line for the light-load line under the maximum load day, and the aim of reducing the line loss is achieved by changing the operation mode of the line for the heavy-load line under the maximum load day through the optimal switching of the switch.
The above-described embodiment is a preferred embodiment of the present invention, and is not intended to limit the present invention in any way, and other variations and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (7)
1. A loss reduction method for automatic reconstruction of a power distribution network is characterized by comprising the following steps:
s1: detecting network data of the power distribution network, and sorting and encoding the data;
s2: calculating the line loss on each branch in the power distribution network by using a forward-backward substitution method;
s3: performing 0/1 coding on each section of the current network, wherein 0 represents that the line is open, and 1 represents that the line is closed, and then constructing a target function;
s4: automatically optimizing and adjusting the operation mode of the power distribution network by utilizing a particle swarm optimization algorithm;
s5: calculating the line loss and the distribution transformation loss after the network reconstruction through a forward-backward substitution method;
s6: comparing the line loss and the distribution transformer loss before and after the reconstruction of the whole looped network switch, and judging whether the adjusted network structure reduces the line loss or not;
in step S4, the specific steps of automatically optimizing and adjusting the operation mode of the power distribution network by using the particle swarm optimization algorithm are as follows:
s4.1: setting initial parameters;
s4.2: forming a topological structure, and calculating the line loss and the total distribution transformer loss by a forward-backward substitution method;
s4.3: evaluating and sequencing the fitness of the particles;
s4.4: updating global extrema according to particle fitness
And individual extremum
All the particle states are then updated, wherein,the global extremum is:
where t represents the number of iterations, m represents the dimension,
representing the optimal position found by the whole particle swarm in the t-th iteration;
the individual extrema are:
in the formula (I), the compound is shown in the specification,
representing the optimal position found by the t iteration particle i;
the updated particle positions are:
wherein K is a convergence factor, r 1 And r 2 Is distributed in [0,1 ]]A random number in between; c. C 1 And c 2 As a learning factor, V i t Representing the velocity, V, of the ith particle at the t-th iteration i t+1 Represents the velocity of the ith particle at the t +1 th iteration, J being a constant;
when the iteration is carried out for t times, the j-th optimal solution searched by the particle swarm is obtained;
s4.5: and judging whether the fitness function termination condition is met, if so, finishing network optimization adjustment, and if not, repeating the steps S4.2-S4.5 until the fitness function termination condition is met.
2. The loss reduction method for automatically reconstructing a power distribution network according to claim 1, wherein the network data collected in step S1 includes: the number of lines, the number of nodes, the number of switching stations, the number of substations and the voltage, current and power are coded by digital coding.
3. The loss reduction method for automatically reconstructing the power distribution network according to claim 1, wherein in the step S2, the line loss of each branch in the power distribution network is calculated by using a forward-backward substitution method in a manner that:
s2.1: inputting network data, searching network nodes and forming a hierarchical relationship;
s2.2: initializing the node voltage:
U i ＝1(i＝0,1,2,3,…,n)
in the formula of U 0 The voltage of the node is 0 point voltage, namely the voltage of the root node of the tree, and n is the number of the nodes;
s2.3: carrying out back substitution calculation: calculating each branch current layer by layer from the last 1 layer of load nodes through the layering of the tree-shaped network;
s2.4: and (3) performing forward calculation: calculating from the root node to the 1 st layer from the 1 st layer to the 1 st layer, and calculating the voltage value of each node layer by layer;
s2.6: according to the formula:
determine whether to converge, wherein U i (t) node Voltage, U, derived from the present Forward calculation i And (t-1) calculating the node voltage obtained by the previous forward calculation, wherein zeta represents the calculation accuracy, if convergence occurs, the line loss on each branch is calculated, and if not, the steps S2.3-S2.5 are repeated until convergence occurs.
4. The loss reduction method for automatically reconstructing the power distribution network according to claim 1, wherein in the step S3, an objective function is constructed to be the minimum of overall loss, including line loss and distribution loss, and the objective function can be expressed as:
wherein, beta is a load factor,
is the load power factor, S N Rated capacity of the transformer, P 0 Representing the power of the root node of the tree, k representing the child node, P k Denotes the power of the child node, l denotes the number of branches, R l Denotes the resistance of the l-th branch, P' l1,2 Representing the active power, Q 'between node 1 and node 2 on the branch' l1,2 Representing reactive power, U ', between node 1 and node 2 on the branch' l1,2 Representing the voltage between node 1 and node 2 on the branch.
5. The loss reduction method for automatically reconstructing the power distribution network according to claim 1, wherein in the step S4.1, the set initial parameters include: an algebra t =0 and an initial scale N are set 0 And setting control parameters and network topology data to form an initial population and an initial particle speed.
6. The loss reduction method for automatically reconstructing the power distribution network according to claim 1, wherein in the step S4.3, the particle fitness is: forward-pushing and backward-replacing the load flow calculation on the network state corresponding to each particle, and taking the corresponding network loss value as an example fitness:
7. the loss reduction method for automatically reconstructing a power distribution network according to claim 1, wherein in step S6, the line loss and the distribution transformation loss before and after the reconstruction of the whole ring network switch are compared, and whether the adjusted network structure reduces the line loss is determined as follows:
and (4) comparing the line loss value obtained in the step (S2) with the line loss value obtained in the step (S5), judging whether the optimized and adjusted network topology structure reduces the line loss, if so, recording the network load state and the current network optimization mode, and if not, returning to the step (S3) to optimize and adjust the network again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110925676.1A CN113673065B (en) | 2021-08-12 | 2021-08-12 | Loss reduction method for automatic reconstruction of power distribution network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110925676.1A CN113673065B (en) | 2021-08-12 | 2021-08-12 | Loss reduction method for automatic reconstruction of power distribution network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113673065A CN113673065A (en) | 2021-11-19 |
| CN113673065B true CN113673065B (en) | 2022-11-18 |
Family
ID=78542540
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110925676.1A Active CN113673065B (en) | 2021-08-12 | 2021-08-12 | Loss reduction method for automatic reconstruction of power distribution network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113673065B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116187723B (en) * | 2023-04-26 | 2023-07-25 | 佰聆数据股份有限公司 | Resource scheduling method and device applied to distribution line loss reduction scene |
| CN117518922A (en) * | 2023-11-21 | 2024-02-06 | 潮州市科莱曼科技有限公司 | Automatic flushing control system of intelligent closestool |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107294081A (en) * | 2016-04-05 | 2017-10-24 | 国家电网公司 | The correlation of line loss per unit influence factor determines method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103474991B (en) * | 2013-09-18 | 2015-11-18 | 国电南瑞科技股份有限公司 | Based on the power distribution network global optimization dispatching method of time scale |
| CN109217284A (en) * | 2017-07-05 | 2019-01-15 | 南京理工大学 | A kind of reconstruction method of power distribution network based on immune binary particle swarm algorithm |
| CN109449994B (en) * | 2018-12-29 | 2021-06-08 | 合肥工业大学 | Power regulation and control method for active power distribution network comprising flexible interconnection device |
| CN110086153A (en) * | 2019-04-15 | 2019-08-02 | 东南大学 | A kind of active power distribution network failure afterload based on intelligent granule colony optimization algorithm turns for method |
-
2021
- 2021-08-12 CN CN202110925676.1A patent/CN113673065B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107294081A (en) * | 2016-04-05 | 2017-10-24 | 国家电网公司 | The correlation of line loss per unit influence factor determines method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113673065A (en) | 2021-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107516170B (en) | Difference self-healing control method based on equipment failure probability and power grid operation risk | |
| CN105741193B (en) | The multiple target Distribution system method of meter and distributed generation resource and negative rules | |
| CN113673065B (en) | Loss reduction method for automatic reconstruction of power distribution network | |
| CN108242807A (en) | A kind of reconstruction method of power distribution network containing photo-voltaic power supply for considering multidimensional security constraint | |
| CN110783925A (en) | Two-stage deployment power grid regulation and control model center control system and method | |
| CN108155649A (en) | A kind of consideration probabilistic distribution network structure Fuzzy Programmings of DG | |
| CN113890015A (en) | Power distribution network dynamic reconstruction method based on improved fuzzy C-means clustering algorithm | |
| CN114069687B (en) | Distributed photovoltaic planning method considering reactive power regulation effect of inverter | |
| CN114742283A (en) | Medium-voltage line loop closing current estimation and risk assessment method based on deep learning network | |
| CN103279661B (en) | Substation capacity Optimal Configuration Method based on Hybrid quantum inspired evolution algorithm | |
| CN113887141A (en) | Micro-grid group operation strategy evolution method based on federal learning | |
| Meng et al. | A new PV generation power prediction model based on GA-BP neural network with artificial classification of history day | |
| CN111832836B (en) | Power distribution network reconstruction method and system considering load power utilization characteristics | |
| Song et al. | Medium and long term load forecasting considering the uncertainty of distributed installed capacity of photovoltaic generation | |
| CN105207255B (en) | A kind of power system peak regulation computational methods suitable for wind power output | |
| CN111555293A (en) | Reactive power equipment type selection system and method based on data model analysis | |
| Qi | Operation ControlMethod of Relay Protection in Flexible DC Distribution Network Compatible with Distributed Power Supply. | |
| CN105550947A (en) | Power distribution network reconstruction method | |
| CN112736905B (en) | Regional new energy capacity optimal configuration method and system | |
| CN113612191A (en) | Method and device for rapidly recovering power supply of power distribution network | |
| CN111952977A (en) | Power market coupling model optimization method, device and medium based on power flow | |
| Li et al. | Influence on Stability Analysis in Distributed Smart Grids Using Computer Aimed Digital Decision Trees | |
| KR102542102B1 (en) | Apparatus for reconfiguration of renewable energy distribution network based on deep reinforcement learning and method of the same | |
| CN112446521B (en) | Multi-objective planning method for wind power plant access system considering economy and safety | |
| Salimon et al. | Optimal placement and sizing of capacitor in nigerian radial distribution networks using cuckoo search algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |