CN210835946U - Equipment and system for researching influence of distribution network line elements on line reliability - Google Patents
Equipment and system for researching influence of distribution network line elements on line reliability Download PDFInfo
- Publication number
- CN210835946U CN210835946U CN201921820071.0U CN201921820071U CN210835946U CN 210835946 U CN210835946 U CN 210835946U CN 201921820071 U CN201921820071 U CN 201921820071U CN 210835946 U CN210835946 U CN 210835946U
- Authority
- CN
- China
- Prior art keywords
- heat dissipation
- line
- heat conducting
- influence
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The utility model discloses a join in marriage influence research equipment of net line key element to circuit reliability, offer the box body of a plurality of jacks including the surface, be equipped with the main control board of being connected through the wire electricity with the jack in the box body, the board carries treater, controller and memory on the main control board, still is equipped with the radiator unit who contacts with the treater in the box body. The utility model discloses simple structure, radiating effect are good, can guarantee the stability when GPU moves effectively, and this device can directly utilize current cloud platform to establish the influence research system of joining in marriage net twine way key element to the circuit reliability that deploys simply, the arithmetic speed is fast.
Description
Technical Field
The utility model relates to an artificial intelligence field, concretely relates to join in marriage influence research equipment and system of net twine way key element to line reliability.
Background
At present, a plurality of fault factors of a line are caused, the problem cannot be found in time only by a traditional expert experience method, and the equipment is large in quantity and large in maintenance workload. The user power supply reliability of the power distribution system is an important content of power reliability management, wherein line fault and power failure are used as important measurement standards of the reliability of the power distribution network and are important indexes for evaluating the grid structure of a power supply enterprise. With the enlargement of the distribution network scale, the standardized management and the influence of main line elements are related to the operation of the distribution network. For the reliability analysis of the distribution network, related algorithms, such as a network equivalence method, exist, and the algorithms are used for equating a plurality of elements of the distribution network into one element, so that the consistency of the elements and the fault power failure is analyzed. But considering the complexity of the distribution network, the similar simplified model cannot reflect the reliability of the distribution network and the elements thereof. Currently, with the development of network technology, the industry starts to gradually use Artificial Intelligence (AI for short) to calculate related data, and the implementation of the Artificial Intelligence algorithm is mainly realized by AI software loaded on a GPU.
The AI software is usually deployed on a distributed cloud platform, which has a high computation speed and a low deployment difficulty, but is expensive and limited by the storage of the platform. If the cloud self-service server is deployed, the manufacturing difficulty is small, but the operation speed is not high, and the servers need to be deployed independently aiming at different regions, so that the cloud self-service server is very inconvenient.
The performance stability of the existing heat dissipation device cannot be well guaranteed due to the large calculation amount of the existing GPU in operation. Although some command instructions exist at present, the operation speed of the fan can be forcibly increased, so that the heat dissipation effect of the GPU is improved, the service life of the fan is also shortened, and when the GPU reaches the limit operation instantly, the peak value of the temperature of the GPU also exceeds the limit of the operation speed of the existing fan, the GPU is easily burnt out, and the limitation is large.
SUMMERY OF THE UTILITY MODEL
In order to solve the problem, the utility model provides a net distribution line key element is joined in marriage influence research equipment and system of line reliability that the radiating effect is good, the operation is stable, and concrete scheme is as follows:
the utility model provides a join in marriage influence research equipment of net line element to circuit reliability, includes that the box body of a plurality of jacks is seted up on the surface, be equipped with in the box body with the jack passes through the main control board that the wire electricity is connected, the board carries treater, controller and memory on the main control board, still be equipped with in the box body with the radiator unit of treater contact.
Further, the heat dissipation assembly comprises a heat dissipation box, a water inlet is formed in the top surface of the heat dissipation box, and the water inlet is connected with a water pump through a pipeline; the bottom surface of the heat dissipation box is provided with a water outlet, and the water outlet is connected with a radiator through a pipeline; the heat dissipation box and the water pump are communicated through a pipeline to form a closed water path, and cooling liquid is arranged in the water path.
Further, the radiator comprises a mosquito-repellent incense-shaped radiating disc and a radiating fan, wherein the mosquito-repellent incense-shaped radiating disc is formed by bending a pipeline between the water outlet and the water pump, the back of the radiating fan is in contact with the mosquito-repellent incense-shaped radiating disc, and a radiating port corresponding to the radiating fan is formed in the box body.
Furthermore, a plurality of heat conducting rods are arranged in the heat dissipation box, one ends of the heat conducting rods are connected to the side wall, connected with the processor, of the heat dissipation box through heat conducting seats, and heat conducting materials are filled between the heat conducting seats and the heat dissipation box.
Furthermore, a plurality of heat conduction fins which are symmetrically arranged are arranged on the heat conduction rod, and the heat conduction fins are vertically arranged relative to the water outlet.
Further, the heat conduction fins on the adjacent heat conduction rods are arranged in a staggered mode.
Further, a heat conduction material is filled between the heat dissipation box and the processor.
The system further comprises a plurality of local servers, each local server is in communication connection with a device for researching the influence of the distribution network line elements on the line reliability, and the device for researching the influence of the distribution network line elements on the line reliability is connected with a distributed cloud platform through a TCP/IP protocol.
The beneficial effects of the utility model reside in that: the utility model provides a pair of join in marriage influence research equipment of net twine way key element to line reliability, simple structure, radiating effect are good, can guarantee the stability when GPU moves effectively, and this device can directly utilize current cloud platform to establish the influence research system of joining in marriage net twine way key element to line reliability that the deployment is simple, the arithmetic speed is fast.
Drawings
FIG. 1 is a schematic structural diagram of the present invention,
FIG. 2 is a schematic view of the structure of the heat dissipation assembly of the present invention,
FIG. 3 is a schematic view of the internal structure of the heat dissipation case of the present invention,
fig. 4, a schematic diagram of the connection of the system of the present invention,
FIG. 5, a decision tree model feature weight graph,
fig. 6, the decision tree results,
FIG. 7 shows average annual average power outage times,
fig. 8, average total line length,
fig. 9 average total on-column transformer capacity.
Figure number and name: 1. jack, 2, box body, 3, heat dissipation box, 4, water inlet, 5, water pump, 6, delivery port, 7, radiator, 8, heat conduction pole, 9, heat conduction seat, 10, heat conduction fin 11, local server, 12, distributed cloud platform.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and examples, which are provided only for explaining the present invention and are not intended to limit the scope of the present invention.
As shown in a combined figure 1, the research equipment for the influence of distribution network line elements on line reliability comprises a box body 2, wherein the surface of the box body 2 is provided with a plurality of jacks 1, a main control board electrically connected with the jacks 1 through conducting wires is arranged in the box body 2, a processor (the model is GeForce 9400GT), a controller (the model is SGM6230) and a memory (the model is SGM6230) are arranged on the main control board, and a heat dissipation assembly in contact with the processor is further arranged in the box body 2.
As shown in fig. 2, the heat dissipation assembly comprises a heat dissipation box 3, a water inlet 4 is formed in the top surface of the heat dissipation box 3, and the water inlet 4 is connected with a water pump 5 through a pipeline; a water outlet 6 is formed in the bottom surface of the heat dissipation box 3, and the water outlet 6 is connected with a radiator 7 through a pipeline; the heat dissipation box 3 is communicated with the water pump 5 through a pipeline to form a closed water path, and cooling liquid is arranged in the water path. The radiator 7 comprises a mosquito-repellent incense-shaped radiating disc and a radiating fan (not shown in the attached drawing), wherein the mosquito-repellent incense-shaped radiating disc is formed by bending a pipeline between the water outlet 6 and the water pump 5, the mosquito-repellent incense-shaped radiating disc is in contact with the back of the radiating fan, and a radiating port corresponding to the radiating fan is formed in the box body 2.
Referring to fig. 3, a plurality of heat conducting rods 8 are arranged in the heat dissipation box 3, one ends of the heat conducting rods 8 are connected to the side wall of the heat dissipation box 3 connected with the processor through heat conducting seats 9, and heat conducting materials are filled between the heat conducting seats 9 and the heat dissipation box 3. The heat conducting rod 8 is provided with a plurality of heat conducting fins 10 which are symmetrically arranged, and the heat conducting fins 10 are vertically arranged relative to the water outlet 6. The heat-conducting fins 10 of adjacent heat-conducting bars 8 are arranged offset from each other. And a heat conduction material is filled between the heat dissipation box 3 and the processor.
Its heat dissipation of current GeForce 9400GT can only rely on the fin that has the fold, the radiating effect is not good, this embodiment radiator unit's structure, on the basis that utilizes the circulating coolant liquid to take away the GPU temperature, set up a plurality of heat conduction rods of taking heat conduction fin in the heat dissipation box, increase and refrigerated area of contact, further improve its radiating effect, heat conduction fin staggered arrangement simultaneously, the effect that makes the coolant liquid take away the heat is better, thereby avoid the GPU chip to cause the damage because the too big heat gathering that produces of operand.
As shown in fig. 4, a system for researching influence of distribution network line elements on line reliability includes a plurality of local servers 11, each local server 11 is connected to the above-mentioned equipment for researching influence of distribution network line elements on line reliability in a communication manner, and the equipment for researching influence of distribution network line elements on line reliability is connected to a distributed cloud platform 12 through a TCP/IP protocol.
The research equipment can realize local operation and independently transmit data with a local server; on the other hand, when certain computation with excessive computation amount is processed, resources of the distributed cloud platform are called to perform subsequent computation, and meanwhile, certain processing results can be uploaded to the distributed cloud platform to be called by other research equipment.
A large amount of equipment ledgers and operation and maintenance data are accumulated in the distribution network line management process, the data are reasonably used, the influence rule of each element of the line on the line reliability is mined, and the method has important significance for scientific and reasonable distribution network fault emergency repair management, timely emergency repair of distribution network faults and improvement of power supply service quality. A prediction model between line attributes and the number of times of fault power failure is established by taking a 10kV large feeder line as a basic research object, key elements influencing the fault power failure of the line are found out, and then sample feeder lines are clustered according to line attribute data through a Kmeans algorithm, and attribute characteristics of samples in each cluster are analyzed, so that the prediction model is verified.
The algorithm principle is as follows:
a decision tree (decision tree) is a simple but widely used model, which is a hierarchical structure composed of nodes and directed edges. In the decision tree, three types of nodes are included: the system comprises a root node, an internal node and leaf nodes, wherein each leaf node is assigned a class label, and non-terminal nodes (including the root node and the internal node) comprise attribute test conditions for separating records with different characteristics.
The purpose of the decision tree model is to generate a branch tree with strong generalization capability, i.e. strong processing capability. How to select the optimal division attribute becomes a difficulty of a decision tree model, and generally, as the division process is continuously carried out, samples contained in the branch nodes can be obtained and belong to the same category as much as possible, namely, the purity of the nodes is higher and higher.
"information entropy" is a most commonly used index for measuring the purity of a sample set, and is defined as formula (1). In the formula: d is the set of samples, pkIs the proportion of the kth sample.
The "information gain" obtained by dividing the sample set D by the attribute a is as shown in formula (2). In the formula: dvThe v-th branch node contains all the values of the attribute a in D as avThe sample of (1).
Utilize the utility model provides a system carries out model establishment and result analysis:
1 model building
Historical failure power failure data of 10kV large feeders of companies in a certain city of a national grid are used as samples, and line self attributes and equipment attributes are used as sample characteristics, wherein the sample characteristics comprise 10 characteristics such as erection mode, overhead line length, cable line length, total line length, power supply area, branch line number, number and total capacity of transformers on columns, number and total capacity of distribution transformers and total capacity. Due to the fact that different lines are different in commissioning time, time factors can be ignored by adopting line power failure statistics, and the sample measurement is unbalanced. Through line power failure times statistics in all years, the annual average power failure times of the line are calculated by combining the line commissioning time and serve as a standard for measuring the reliability of the line.
The feature weights of the lines are visualized by building a prediction model through a decision tree as shown in fig. 5. As is evident from fig. 5, the total length of the line is highly weighted on the number of line outages, followed by the total capacity of the pole transformer.
Post-pruning the tree model and the final decision tree is shown in fig. 6. The actual value and the predicted value are an actual value and a predicted value of the average annual power failure times of the sample respectively. It can be seen that after pruning, the line power failure can be predicted only by the total line length and the total capacity of the transformer on the column. The average value of the annual fault power failure prediction times of the sample is 3.57 times, the average deviation is 0.29 times, the deviation rate is 7.51%, and the overall prediction accuracy is high. When the total length of the line is more than 19.451km and the total capacity of the pole transformer is more than 7122.5kVA, the average power failure times of the line are higher.
For the sample, a Kmeans clustering model is established based on the selected 10 line characteristics, the sample data is divided into 6 classes through contour coefficient selection, and the average value of the main characteristics of each class is shown in table 1.
TABLE 1 mean values of the characteristics of the clusters
2 analysis of results
FIG. 7 shows the average annual average power outage for each cluster type of sample data. As can be seen from the figure, the average number of actual annual blackouts for the cluster 1 sample line reaches 7.05, which is much higher than for the other clusters. The next year, the number of power failures per year of cluster 4 is 4.19.
Fig. 8 shows the average total line length for each cluster. As can be seen from the figure, the total length of the lines of both cluster 1 and cluster 4 is above 25km, which is higher than the total length of the lines of the other clusters. And through correlation analysis with the table 4, the length of each cluster of lines has a strong positive correlation with the failure power failure frequency, and the longer the line is, the higher the power failure frequency is.
Fig. 9 shows the average total on-column transformer capacity for each cluster. It can be seen from the figure that the total on-column transformer capacity of both cluster 1 and cluster 4 is above 6000kVA, which is higher than that of the other clusters. The positive correlation between the capacity of the transformers on the columns of each cluster of the circuit and the failure power failure frequency is shown, and the higher the total capacity of the transformers on the columns of the circuit is, the higher the failure power failure frequency is.
The cluster analysis results show that the distribution of the total line length and the total capacity of the pole transformer among clusters is basically the same as the distribution of the average power failure times of the lines, and the reliability of the decision tree model conclusion is further verified.
And (4) conclusion:
by establishing a decision tree model between the distribution network line attribute and the fault power failure frequency, the main factors influencing the line fault power failure are the total line length and the total capacity of the transformer on the column on the line, and the total line length and the total capacity are in positive correlation with the line power failure frequency. Meanwhile, the Kmeans algorithm is adopted to perform characteristic clustering on the samples, and the correctness of the decision tree conclusion is further verified.
In summary, the preferred embodiments of the present invention are only examples, and the scope of the present invention is not limited thereto, and all equivalent changes and modifications made in accordance with the scope of the present invention and the contents of the specification are within the scope covered by the present invention.
Claims (5)
1. The utility model provides a join in marriage net line element to influence research equipment of line reliability, includes box body (2) that a plurality of jacks (1) were seted up on the surface, its characterized in that: a main control board electrically connected with the jack (1) through a wire is arranged in the box body (2), a processor, a controller and a memory are arranged on the main control board, and a heat dissipation assembly in contact with the processor is also arranged in the box body (2);
the heat dissipation assembly comprises a heat dissipation box (3), a water inlet (4) is formed in the top surface of the heat dissipation box (3), and the water inlet (4) is connected with a water pump (5) through a pipeline; a water outlet (6) is formed in the bottom surface of the heat dissipation box (3), and the water outlet (6) is connected with a radiator (7) through a pipeline; the heat dissipation box (3) is communicated with the water pump (5) through a pipeline to form a closed water path, and cooling liquid is arranged in the water path;
the radiator (7) comprises a mosquito-repellent incense-shaped radiating disc and a radiating fan, wherein the mosquito-repellent incense-shaped radiating disc is formed by bending a pipeline between the water outlet (6) and the water pump (5), the mosquito-repellent incense-shaped radiating disc is in contact with the back of the radiating fan, and a radiating port corresponding to the radiating fan is formed in the box body (2);
the heat dissipation box is characterized in that a plurality of heat conducting rods (8) are arranged in the heat dissipation box (3), one ends of the heat conducting rods (8) are connected to the side wall, connected with the processor, of the heat dissipation box (3) through heat conducting seats (9), and heat conducting materials are filled between the heat conducting seats (9) and the heat dissipation box (3).
2. The device for researching influence of distribution network line elements on line reliability, according to claim 1, is characterized in that: the heat conducting rod (8) is provided with a plurality of heat conducting fins (10) which are symmetrically arranged, and the heat conducting fins (10) are vertically arranged relative to the water outlet (6).
3. The device for researching influence of distribution network line elements on line reliability, according to claim 2, is characterized in that: the heat conducting fins (10) on the adjacent heat conducting rods (8) are arranged in a staggered mode.
4. The device for researching influence of distribution network line elements on line reliability, according to claim 1, is characterized in that: and a heat conduction material is filled between the heat dissipation box (3) and the processor.
5. The utility model provides a join in marriage influence research system of net line element to line reliability which characterized in that: the system comprises a plurality of local servers (11), wherein each local server (11) is connected with a device for researching the influence of the distribution network line elements on the line reliability according to any one of claims 1-4 in a communication mode, and the device for researching the influence of the distribution network line elements on the line reliability is connected with a distributed cloud platform (12) through a TCP/IP protocol.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201921820071.0U CN210835946U (en) | 2019-10-25 | 2019-10-25 | Equipment and system for researching influence of distribution network line elements on line reliability |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201921820071.0U CN210835946U (en) | 2019-10-25 | 2019-10-25 | Equipment and system for researching influence of distribution network line elements on line reliability |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN210835946U true CN210835946U (en) | 2020-06-23 |
Family
ID=71252418
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201921820071.0U Expired - Fee Related CN210835946U (en) | 2019-10-25 | 2019-10-25 | Equipment and system for researching influence of distribution network line elements on line reliability |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN210835946U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111768109A (en) * | 2020-07-02 | 2020-10-13 | 广东电网有限责任公司 | Reliability early warning method and system for power electronic medium-voltage distribution network and terminal equipment |
-
2019
- 2019-10-25 CN CN201921820071.0U patent/CN210835946U/en not_active Expired - Fee Related
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111768109A (en) * | 2020-07-02 | 2020-10-13 | 广东电网有限责任公司 | Reliability early warning method and system for power electronic medium-voltage distribution network and terminal equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109687458B (en) | Grid planning method considering regional distribution network risk bearing capacity difference | |
| CN110210659B (en) | Power distribution network planning method considering reliability constraint | |
| WO2013174145A1 (en) | Large-scale wind power grid-integration reactive voltage optimization method based on improved artificial fish swarm hybrid optimization algorithm | |
| CN109657959A (en) | A kind of distribution network planning calculation and analysis methods containing multivariate data | |
| WO2019154135A1 (en) | Microgrid reliability evaluation method based on network survivability | |
| CN111428376A (en) | CIM (common information model) -based power distribution network feeder contact mode identification method and system | |
| CN110188972B (en) | 10kV power distribution network non-private line customer access method | |
| CN210835946U (en) | Equipment and system for researching influence of distribution network line elements on line reliability | |
| CN112924817A (en) | Power distribution network fault accurate positioning method based on deep neural network | |
| CN112380676A (en) | Multi-energy system digital twin data stream modeling and compressing method | |
| CN113904322A (en) | Low-voltage distribution network topology generation method based on current and voltage | |
| CN114050578B (en) | Power distribution network online load flow calculation method, device, equipment and medium | |
| CN111222097B (en) | Reliability evaluation and sensitivity analysis method for photovoltaic direct-current boosting and collecting system | |
| CN114186858A (en) | Method and system for evaluating importance of energy storage planning node of power distribution network | |
| CN209342847U (en) | Distribution line fault outage prediction meanss and system based on weather and XBoost algorithm | |
| Pahwa et al. | A complex networks approach for sizing and siting of distributed generators in the distribution system | |
| CN205561786U (en) | Super high voltage transmission wire arc intelligent measuring device that hangs down | |
| CN104463365A (en) | Reactive voltage optimization analysis and assessment method based on distribution network automation | |
| CN109887260A (en) | Platform area electric energy meter network topology structure method for splitting | |
| CN108258687A (en) | A kind of method of determining Distributed Generation in Distribution System on-position | |
| CN111553040B (en) | Power grid topology analysis high-performance computing method and device based on GPU acceleration | |
| CN115186975A (en) | Power grid line loss management method and system based on service center | |
| CN105488621B (en) | The analysis of power grid use for laboratory wastage reducing and energy saving and efficiency evaluating system | |
| CN104820954B (en) | The system of selection of 110kV substation locations and system | |
| CN115642597B (en) | Distributed photovoltaic bearing capacity calculation method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200623 Termination date: 20201025 |