CN105651460A - Prediction method and apparatus for fuel tank sealing index of distribution network transformer - Google Patents

Abstract

The invention discloses an automatic device of a distribution network and especially relates to a prediction method and apparatus for a fuel tank sealing index of a distribution network transformer. The prediction apparatus is characterized in that the apparatus comprises a signal acquisition module installed on a distribution network transformer; the output terminal of the signal acquisition module is connected to an input terminal of an A/D converter by a signal conversion circuit; the output terminal of the A/D converter is connected to an input terminal of a central processing unit (CPU); a serial port of the CPU is connected with a 4G communication transmission module and a man-machine interaction information display module; and an ATN1 terminal of the 4G communication transmission module carries out communication with a remote scheduling center by an antenna. According to the invention, a fuel tank sealing index of a distribution network transformer can be predicted, thereby preventing a major accident of a power grid, improving the power quality, and enhancing the power usage reliability. Meanwhile, a real-time requirement is met during the prediction process; the data collection and processing efficiency is improved; and the prediction speed and precision of the fuel tank sealing index of the distribution network transformer.

Description

A kind of network distribution transformer fuel tank sealing index forecasting method and device

Technical field

The present invention relates to a kind of distribution network automatic gear, particularly a kind of network distribution transformer fuel tank sealing index forecasting method and device, belong to technical field of power systems.

Background technology

Substation transformer high pressure is one of main and the most important equipment of distribution network, and transformer can generate heat owing to flowing through electric current in winding in operational process, for ensureing that winding temperature rise is no more than limit value, will be cooled by transformer by the transformer oil in fuel tank. If oil tank of transformer tightness reduces, transformer oil being caused to leak, and the leakage of transformer oil causes oil amount minimizing and the circulation of oil in fuel tank to be deteriorated, this will cause transformer winding temperature out of hand again further, finally causes transformer fault. Transformer generation fault will cause large area blackout, seriously affect supplying power allocation reliability and electric power netting safe running. Therefore, accurately grasping network distribution transformer fuel tank sealing index, to Regulation, department is most important. Current traditional method mainly adopts that tradition experimental formula calculates or makes regular check on, the method such as replacing, the main problem of existence be can not Timeliness coverage fault, judged result error is bigger.

Summary of the invention

It is an object of the present invention to provide a kind of network distribution transformer fuel tank sealing index forecasting method, the network distribution transformer fuel tank sealing exponential forecasting model that it is applicable to practical implementation by setting up, effectively to solve the accurately predicting that network distribution transformer fuel tank is sealed index, reach the object finding fault constantly.

It is a further object to provide a kind of network distribution transformer fuel tank sealing exponential forecasting device, its input is extracted simple, predicts that process requirement of real time, prediction accuracy height, time of response are short.

In order to achieve the above object, the present invention is achieved through the following technical solutions:

A kind of network distribution transformer fuel tank sealing index forecasting method, its step is as follows:

Step one, set up network distribution transformer fuel tank sealing index evolution system state vector: in Fixed Time Interval measuring transformer oil pressure, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture, in a series of moment t₁,t₂,...,t_n(n is natural number, n=1,2 ...) obtain transformer oil pressure, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture form a series of state vectors:

$\left\{\begin{array}{c}{X}_{t_1}={(x_{11},x_{12},x_{13},x_{14},x_{15})}^T\\ X_{t_2}={(x_{21},x_{22},x_{23},x_{24},x_{25})}^T\\ ...\\ X_{t_n}={(x_{n1},x_{n2},x_{n3},x_{n4},x_{n5})}^T\end{array}\right.$

Step 2, calculating initial weight vector: given t₁MomentCorresponding connection weight vector $W_{t_1}={(w_{11},w_{12},w_{13},w_{14},w_{15})}^T,$ It is initial weight vector;

Step 3, correction weight vector: after the state vector in any one moment inputs, weight vector adjustment amount is:

$\mathrm{\Δ}{W}_{t_n}=(\mathrm{\η}-1)\sqrt{{(d-{W_{t_n}}^T{X}_{t_n})}^2}{X}_{t_n}$

Its d is the expected value of fusible cut-out load factor predictor, and �� is accelerator coefficient;

The actual prediction value making oil tank of transformer seal index is

$y_{t_n}={W_{t_n}}^T{X}_{t_n}$

Definition miscalculationFor the difference of expected value and actual prediction value, have:

${\mathrm{\ϵ}}_{t_n}=d-y_{t_n}$

Then weighed value adjusting equation is:

$\mathrm{\Δ}{W}_{t_n}=(\mathrm{\η}-1)\sqrt{\frac{1}{{\left|\right|X_{t_n}\left|\right|}^2}{\mathrm{\ϵ}}_{t_n}{X}_{t_n}}$

Step 4, calculating weight vector end value W₀: definition error rateFor:IfWithSymbol is contrary, illustrates in the calculation,Absolute value be monotonic decreasing, nowAlways constantly close to d; If t_nShi Keyou:WithSymbol contrary andBe less than set-point, then now correspondingIt is weight vector end value W₀;

Step 5, calculating network distribution transformer fuel tank sealing exponential forecasting value: for any instant t_n+1, according to W₀, have:

$y_{t_{n+1}}={W_{t_{n+1}}}^T{X}_{t_{n+1}}$

It is network distribution transformer fuel tank sealing exponential forecasting value.

A kind of network distribution transformer fuel tank sealing exponential forecasting device, its structure is as follows: the signal acquisition module being arranged on network distribution transformer, the output terminal of signal acquisition module is connected to A/D converter input terminus after changing circuit respectively through signal connects, A/D converter adopts TLC2543 serial a/d transmodulator, the output terminal of A/D converter is connected with the input terminus of cpu central processing unit, cpu central processing unit adopts model to be the micro-chip of STM32F103RBT6, the serial ports of cpu central processing unit is connected with 4G communications module and human-machine interactive information display module, 4G communications module carries out communication by antenna and remote dispatching center.

4G communications module adopts the LTE module of ME3760 model, and human-machine interactive information display module adopts the liquid-crystal display module of HG1286402C model.

The output terminal of above-mentioned signal acquisition module is connected to the input terminus AIN0 of A/D converter TLC2543 after changing circuit respectively through signal, AIN1, AIN2, AIN3, AIN4, the output terminal EOC of A/D converter TLC2543, I/O, IN, OUT, CS is connected respectively to the PC0 of micro-chip STM32F103RBT6 chip, PC1, PC3, PC4, PC5 pin, the PA1 of micro-chip STM32F103RBT6 chip, PA2, PA3, PA4, PA5, PA6, PA7, the D0 of PC7 and human-machine interactive information display module, D1, D2, D3, D4, D5, D6, D7 connects, the PC8 of micro-chip STM32F103RBT6 chip, PC9, PC10, PC11, the RS of PC12 and human-machine interactive information display module, R/W, CS1, CS2, EN is connected, the PB10 of micro-chip STM32F103RBT6 chip, the DATA of PB11 pin and 4G communications module ME3760, DATA1 end is connected, the ATN1 end of 4G communications module carries out communication by antenna and remote dispatching center.

Above-mentioned signal acquisition module is made up of pressure transmitter, current transformer, voltage transformer, humidity sensor and temperature sensor.

Above-mentioned pressure transmitter adopts PTG501 model oil pressure sensor, current transformer to adopt DHC03B model, voltage transformer to adopt DH51D6V0.4B model, humidity sensor to adopt STYB3100111A50 model, temperature sensor to adopt HE-200 model infrared temperature sensor.

The present invention compared with prior art has following outstanding useful effect:

A kind of network distribution transformer fuel tank sealing exponential forecasting device of the present invention, propose to utilize the oil pressure of the direct measuring transformer of signal acquisition module, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture to be input, and finally utilize A/D converter, cpu central processing unit, human-machine interactive information display module and 4G communications module to realize network distribution transformer fuel tank sealing exponential forecasting.The error that this kind of structure causes when avoiding traditional method Modling model and choose parameter, and there is input and extract simple, tolerance range height, accuracy is good, the feature that forecasting efficiency is high, for the safe operation of network distribution transformer provides effective guarantee.

Network distribution transformer fuel tank can accurately be sealed index and make prediction by the present invention, electrical network Serious Accident can be prevented, improve power quality, improve electricity consumption reliability, predict process requirement of real time simultaneously, improve the efficiency of data gathering and process, it is to increase the speed of network distribution transformer fuel tank sealing exponential forecasting and precision, it is achieved that with degree of precision and the advantage of shorter time of response, network distribution transformer fuel tank is sealed index and predict.

Accompanying drawing explanation

Fig. 1 is network distribution transformer fuel tank sealing exponential forecasting schema.

Fig. 2 is the structure schematic block diagram of the present invention.

Fig. 3 is the circuit connection diagram of the present invention.

Fig. 4 is the figure that predicts the outcome of the present invention.

In figure, 1, signal acquisition module, 2, A/D converter, 3, cpu central processing unit, 4, human-machine interactive information display module, 5,4G communications module.

Embodiment:

Below in conjunction with embodiment, the present invention is described in further detail. Following examples are only several specific embodiments of the present invention, but the design of the present invention design is not limited thereto, and all changes utilizing this design that the present invention carries out unsubstantiality, all should within protection domain of the present invention.

As shown in Figure 1, a kind of network distribution transformer fuel tank sealing index forecasting method of the present invention, its step is as follows:

Step one, set up network distribution transformer fuel tank sealing index evolution system state vector:

In the present embodiment, measuring transformer oil pressure, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture within the time period at the intervals such as 20, then obtain one the 5 dimension time series that take off data is formed:

$\left\{\begin{array}{c}{X}_{t_1}={(x_{11},x_{12},x_{13},x_{14},x_{15})}^T\\ X_{t_2}={(x_{21},x_{22},x_{23},x_{24},x_{25})}^T\\ ...\\ X_{t_{20}}={(x_{201},x_{202},x_{203},x_{204},x_{205})}^T\end{array}\right.---\left(8\right)$

Step 2, calculating initial weight vector:

Given t₁MomentCorresponding connection weight vector $W_{t_1}={\left(\mathrm{0.111,0.211,0.312,0.1324,0.2457}\right)}^T,$ For initial weight vector;

Step 3, correction weight vector:

Given transformer mailbox sealing index expected value d=0.6898, byCalculateHave:

${\mathrm{\ϵ}}_{t_n}=d-y_{t_n}---\left(9\right)$

Make accelerator coefficient ��=1.675, then modified weight amount is:

$\mathrm{\Δ}{W}_{t_n}=(\mathrm{\η}-1)\sqrt{\frac{1}{{\left|\right|X_{t_n}\left|\right|}^2}{\mathrm{\ϵ}}_{t_n}{X}_{t_n}}---\left(10\right)$

Step 4, calculating weight vector end value W₀:

Calculate t_nThe error rate in moment

$\mathrm{\Δ}{\mathrm{\ϵ}}_{t_n}=-\mathrm{\Δ}{W_{t_n}}^T{X}_{t_n}---\left(7\right)$

If t_nShi Keyou:WithSymbol contrary andThen nowIt is network distribution transformer fuel tank sealing exponential forecasting value;

Step 5, calculating network distribution transformer fuel tank sealing exponential forecasting value:

For any instant t_n+1, according to W₀, have:

$y_{t_{n+1}}={W_{t_{n+1}}}^T{X}_{t_{n+1}}$

It is network distribution transformer fuel tank sealing exponential forecasting value.

The part of predictive model predicts the outcome and compares as shown in Figure 4 with measured value, and the error prediction model set up controls substantially within �� 10%.

As shown in Figure 2, a kind of network distribution transformer fuel tank sealing exponential forecasting device of the present invention, comprises signal acquisition module 1, A/D converter 2, cpu central processing unit 3, human-machine interactive information display module 4 and 4G communications module 5; Signal acquisition module is made up of pressure transmitter, current transformer, voltage transformer, humidity sensor and temperature sensor. Its structure is: the signal acquisition module 1 being arranged on network distribution transformer, the output terminal of signal acquisition module 1 is connected to A/D converter 2 input terminus after changing circuit respectively through signal connects, the output terminal of A/D converter 2 is connected with the input terminus of cpu central processing unit 3, the serial ports of cpu central processing unit 3 is connected with 4G communications module 5 and human-machine interactive information display module 4, the output terminal of 4G communications module 5 and long-range upper machine communication.Pressure transmitter adopts PTG501 model oil pressure sensor, current transformer to adopt DHC03B model, voltage transformer to adopt DH51D6V0.4B model, humidity sensor to adopt STYB3100111A50 model, temperature sensor to adopt HE-200 model infrared temperature sensor; A/D converter 2 adopts TLC2543 serial a/d transmodulator; Cpu central processing unit 3 adopts model to be the micro-chip of STM32F103RBT6; 4G communications module 5 adopts the LTE module of ME3760 model; Human-machine interactive information display module 4 adopts the liquid-crystal display module of HG1286402C model.

The output terminal of the pressure transmitter of signal acquisition module 1, current transformer, voltage transformer, humidity sensor, temperature sensor is connected to input terminus AIN0, AIN1, AIN2, AIN3, AIN4 of the TLC2543 of A/D converter 2 after changing circuit respectively through signal, as shown in Figure 3, the output terminal EOC of A/D converter 2TLC2543, I/O, IN, OUT, CS is connected respectively to the PC0 of micro-chip STM32F103RBT6 chip, PC1, PC3, PC4, PC5 pin, the PA1 of micro-chip STM32F103RBT6 chip, PA2, PA3, PA4, PA5, PA6, PA7, the D0 of PC7 and HG1286402C liquid-crystal display module, D1, D2, D3, D4, D5, D6, D7 connects, the PC8 of micro-chip STM32F103RBT6 chip, PC9, PC10, PC11, the RS of PC12 and HG1286402C liquid-crystal display module, R/W, CS1, CS2, EN is connected, the PB10 of STM32F103RBT6 chip, PB11 pin communicates with 4G the DATA of module ME3760, DATA1 end is connected, the ATN1 end of 4G communications module carries out communication by antenna and remote dispatching center.

The principle of work of the present invention is as follows: transformer oil pressure, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture signal are through each sensor, carry out synchronized sampling, maintenance, A/D conversion, after turning into numerary signal, send into the data input port of ARM micro-chip, after network distribution transformer fuel tank sealing exponent data being calculated by micro-chip again, calculation result is undertaken showing by liquid-crystal display module and data is delivered to 4G communications module, for ready with remote dispatching communication. A/D converter is used for converting the simulating signal of signal gathering unit collection to numerary signal; This device uses switch electric capacity Approach by inchmeal technology to complete A/D switching process; Human-machine interactive information display unit is for showing network distribution transformer fuel tank sealing exponential forecasting result; For transmitting, network distribution transformer fuel tank sealing exponential forecasting result is transferred to dispatching center to 4G communications module.

The particular of the present invention is to invention has been detailed description; for a person skilled in the art, the various apparent change it carried out when not deviating from the spirit and scope of the present invention is all within protection scope of the present invention.

Claims (5)

1. a network distribution transformer fuel tank sealing index forecasting method, it is characterised in that step is as follows:

Step one, set up network distribution transformer fuel tank sealing index evolution system state vector: in Fixed Time Interval measuring transformer oil pressure, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture, in a series of moment t₁,t₂,...,t_n(n is natural number, n=1,2 ...) obtain transformer oil pressure, transformer oil temperature, transformer current, transformer voltage, atmospheric moisture form a series of state vectors:

$\left\{\begin{array}{c}{X}_{t_1}={(x_{11},x_{12},x_{13},x_{14},x_{15})}^T\\ X_{t_2}={(x_{21},x_{22},x_{23},x_{24},x_{25})}^T\\ \·\·\·\\ X_{t_n}={(x_{n1},x_{n2},x_{n3},x_{n4},x_{n5})}^T\end{array}\right.$

Step 2, calculating initial weight vector: given t₁MomentCorresponding connection weight vector ${W_t}_1={(w_{11},w_{12},w_{13},w_{14},w_{15})}^T$ , it is initial weight vector;

Step 3, correction weight vector: after the state vector in any one moment inputs, weight vector adjustment amount is:

${\mathrm{\ΔW}}_{t_n}=(\mathrm{\η}-1)\sqrt{{(d-{W_{t_n}}^T{X}_{t_n})}^2}{X}_{t_n}$

Its d is the expected value of fusible cut-out load factor predictor, and �� is accelerator coefficient;

The actual prediction value of fusible cut-out load factor is made to be:

$y_{t_n}={W_{t_n}}^T{X}_{t_n}$

Definition miscalculationFor the difference of expected value and actual prediction value, have:

${\mathrm{\ϵ}}_{t_n}=d-y_{t_n}$

Then weighed value adjusting equation is:

${\mathrm{\ΔW}}_{t_n}=(\mathrm{\η}-1)\sqrt{\frac{1}{{\left|\right|X_{t_n}\left|\right|}^2}{\mathrm{\ϵ}}_{t_n}{X}_{t_n}}$

Step 4, calculating weight vector end value W₀: definition error rateFor:IfWithSymbol is contrary, illustrates in the calculation,Absolute value be monotonic decreasing, nowAlways constantly close to d; If t_nShi Keyou:WithSymbol contrary andBe less than set-point, then now correspondingIt is weight vector end value W₀;

Step 5, calculating network distribution transformer fuel tank sealing exponential forecasting value: for any instant t_n+1, according to W₀, have:

$y_{t_{n+1}}={W_{t_{n+1}}}^T{X}_{t_{n+1}}$

It is network distribution transformer fuel tank sealing exponential forecasting value.

2. network distribution transformer fuel tank sealing exponential forecasting device described in a claim 1, it is characterized in that structure is as follows: the signal acquisition module being arranged on network distribution transformer, the output terminal of signal acquisition module is connected to A/D converter input terminus after changing circuit respectively through signal connects, A/D converter adopts TLC2543 serial a/d transmodulator, the output terminal of A/D converter is connected with the input terminus of cpu central processing unit, cpu central processing unit adopts model to be the micro-chip of STM32F103RBT6, the serial ports of cpu central processing unit is connected with 4G communications module and human-machine interactive information display module, the output terminal of 4G communications module and long-range upper machine communication, 4G communications module adopts the LTE module of ME3760 model, human-machine interactive information display module adopts the liquid-crystal display module of HG1286402C model.

3. a kind of network distribution transformer fuel tank sealing exponential forecasting device according to claim 2, it is characterised in that the output terminal of described signal acquisition module is connected to the input terminus AIN0 of A/D converter TLC2543 after changing circuit respectively through signal, AIN1, AIN2, AIN3, the output terminal EOC of AIN4, A/D converter TLC2543, I/O, IN, OUT, CS is connected respectively to the PC0 of micro-chip STM32F103RBT6 chip, PC1, PC3, PC4, PC5 pin, the PA1 of micro-chip STM32F103RBT6 chip, PA2, PA3, PA4, PA5, PA6, PA7, the D0 of PC7 and human-machine interactive information display module, D1, D2, D3, D4, D5, D6, D7 connects, the PC8 of micro-chip STM32F103RBT6 chip, PC9, PC10, PC11, the RS of PC12 and human-machine interactive information display module, R/W, CS1, CS2, EN is connected, the PB10 of micro-chip STM32F103RBT6 chip, the DATA of PB11 pin and 4G communications module ME3760, DATA1 end is connected, and the ATN1 end of 4G communications module carries out communication by antenna and remote dispatching center.

4. a kind of network distribution transformer fuel tank sealing exponential forecasting device according to claim 2, it is characterised in that described signal acquisition module is made up of pressure transmitter, current transformer, voltage transformer, humidity sensor and temperature sensor.

5. a kind of network distribution transformer fuel tank sealing exponential forecasting device according to claim 4, it is characterised in that described pressure transmitter adopts PTG501 model oil pressure sensor, current transformer to adopt DHC03B model, voltage transformer to adopt DH51D6V0.4B model, humidity sensor to adopt STYB3100111A50 model, temperature sensor to adopt HE-200 model infrared temperature sensor.