CN106197710A - High pressure equipment wireless temperature online detection system based on SVM prediction algorithm - Google Patents

Abstract

The present invention passes through wireless temperature measurement Technology Integration one on-line temperature monitoring, prognoses system, reaches the function being monitored real time temperature, subsequent period temperature is predicted.This system, carries out temperature acquisition initially with temperature sensor, and temperature signal is wirelessly transferred to terminal monitoring equipment；Thereafter, by SVM prediction subsequent period temperature.Finally, temperature monitoring and prediction data are shown, store, report to the police by terminal.This system solves the High-Voltage Insulation problem between switch cubicle temp measuring system high and low pressure side well, need not connect up, and cost is relatively low.Can not having a power failure, ensure safety in the case of, in real time, measure efficiently and accurately, dope temperature, find temperature anomaly, fix a breakdown in time, ensure the reliable and secure stable operation of electrical network.Significant to the high pressure equipment management of current China.

Description

Wireless Temperature Online Monitoring System of High Voltage Equipment Based on Support Vector Machine Prediction Algorithm

【Technical field】

This invention is a high-tech science and technology project, and the subject matter is an online temperature monitoring system with temperature prediction function applied to high-voltage equipment.

【Background technique】

Most of the high-voltage switchgear adopts a closed structure, with poor heat dissipation conditions and large heat accumulation. Under the long-term operation of the equipment under high voltage, high current and full load conditions, the contacts or joints in the switch cabinet are due to aging and excessive contact resistance. Large and hot, as a result, heat accumulation is intensified, and the temperature rise directly threatens the electrical insulation of electrical equipment. The initial performance is that the temperature rises, which in turn causes a short circuit, resulting in a large area of cable burning, resulting in forced shutdown, and cannot resume production in a short time , and even cause a fire, resulting in heavy losses. At this time, it becomes an effective solution to take effective monitoring measures for the temperature rise of the high-voltage switchgear and find out its overheating in time.

At present, the monitoring of high-voltage equipment is mostly based on manual inspections. Hand-held infrared thermometers obtain temperature data in switch cabinets. Accurate temperature data; at the same time, errors in manual operation cannot be completely avoided, and the inspection interval is too long compared to the time when the fault occurs. Therefore, it is urgent to study the online monitoring and early warning system. The online monitoring of the high-voltage cabinet temperature finds abnormalities and the online prediction of the temperature through data mining technology to eliminate hidden dangers in time.

【Content of invention】

The invention integrates a temperature online monitoring system through the wireless temperature measurement technology and combines it with the data mining prediction technology, and performs online prediction of the temperature at the next moment while monitoring the real-time temperature online. The system uses a thermoelectric temperature sensor for temperature collection, and the temperature signal is transmitted to the terminal monitoring equipment through wireless. This method can well solve the problem of high-voltage insulation between the high and low voltage sides of the temperature measurement system of the switchgear, without wiring, and the cost is relatively low. Low. At present, some temperature sensors are powered by high-performance lithium batteries. As a disposable battery to maintain the work of the sensor, its theoretical life can reach 6 years. In actual use, due to the influence of temperature conditions in the working environment, the working life of the battery may More than 5 years, and the cost is not very high. And combined with the current relatively mature prediction technology to predict the temperature online, it can not only provide high-temperature alarms, but also provide technical support for early warnings, so that monitoring personnel can better manage equipment, which is of great significance to the current high-voltage equipment management in my country. The block diagram of the high-voltage equipment wireless temperature online monitoring system based on the support vector machine prediction algorithm is shown in Figure 1.

The whole system is divided into three parts: temperature acquisition transmitter, receiver, host computer control terminal. The temperature acquisition transmitter includes: a temperature sensor, a microprocessor, and a radio frequency transmitter module. The assembled temperature measuring device is placed at the measured point, the temperature sensor converts the temperature signal into a digital signal, the microprocessor packages the data, and the radio frequency module sends the data package to the receiving end. The receiving end includes: a radio frequency receiving module and a microprocessor. Its structure is similar to that of the temperature acquisition transmitter. The microprocessor controls the reception of the radio frequency transceiver module. The radio frequency receiver module receives wireless signals and stores them in registers. The functions of the main control terminal of the upper computer mainly include: communication, display, alarm, prediction, data storage and event recording.

The working process of the system: the sensor installed at the temperature measurement point obtains energy from the high-voltage equipment and converts the temperature signal into a digital signal, processes and packs the collected temperature data through the microprocessor, and transmits the data to the receiving end by using the wireless radio frequency transmitting module , the radio frequency receiving module at the receiving end receives the wireless signal from the same address code, stores the data in the register, extracts the data through the microprocessor, and then uploads it to the upper computer monitoring system through the data bus. The main control terminal of the upper computer receives the data from the receiving end and displays it on the screen, displaying the real-time temperature curve. If the temperature exceeds the limit, the monitoring host will display the alarm information and pop up the alarm icon. At the same time, use data mining prediction technology to predict the data and display it on the word screen. The main program flow chart of the system is shown in Figure 2.

The invention adopts a prediction algorithm based on a support vector machine applied to a wireless temperature measurement system of high-voltage equipment. The support vector machine method is based on the VC dimension theory of statistical learning theory and the principle of minimum structural risk. Basically, it does not involve the definition of probability measurement and the law of large numbers, etc. It combines optimization, kernel, and best promotion ability. The prediction algorithm establishes the objective function for the historical data samples, obtains the regression function by using the support vector machine method, and predicts the data of the next period. The design process of the prediction algorithm based on the support vector machine of the wireless online temperature measurement system is shown in Figure 3.

Compared with the existing scheme, the present invention has the following significant beneficial effects: (1) The data exchange between the upper and lower computers is completed by means of wireless communication, and non-contact monitoring is realized. Not limited by the transmission line, the complexity is reduced; (2) Real-time, effective and accurate measurement of the busbar near the static contact of the switchgear, the outgoing cable connector of the switchgear, and the outdoor high-voltage switch, etc. The temperature of the part can be detected accurately to detect abnormal temperature changes, so as to troubleshoot in time and ensure the reliable, safe and stable operation of the power grid. (3) Use the support vector machine to predict the change of the temperature at the next moment relative to the temperature in the selected similar time period, and add the predicted value of the change to the temperature in the similar time period to obtain the basic item of temperature. It provides a good and stable early warning measure for the safety of high voltage equipment.

【Description of drawings】

Fig. 1 is a block diagram of a wireless temperature online monitoring system for high-voltage equipment based on a support vector machine prediction algorithm described in the patent of the present invention;

Fig. 2 is the main program flow chart of the system described in the patent of the present invention;

Fig. 3 is the prediction algorithm flow based on the support vector machine of the wireless online temperature measurement system of the patent of the present invention;

【detailed description】

The present invention combines the system block diagram shown in FIG. 1 , the main program flow chart shown in FIG. 2 and the prediction algorithm flow chart shown in FIG. 3 as a novel implementation scheme. The steps of the high-voltage equipment wireless temperature online monitoring system based on the support vector machine prediction algorithm are as follows:

(1), initialization parameters;

(2) Serial port address configuration between the upper computer and the receiving end microprocessor;

(3) Calculate the prediction decision function based on the past historical data samples, predict the temperature in the next period, and display it;

(4) The temperature sensor at the temperature acquisition transmitter collects temperature data and converts it into a digital signal;

(5), the temperature acquisition transmitter microprocessor packs the data;

(6), the radio frequency module of the temperature acquisition transmitting end sends the data packet to the receiving end;

(7), the radio frequency receiving module of the receiving end receives the wireless signal from the same address code, and stores the data in the register;

(8), the receiving end microprocessor extracts the data in the register, and transmits the data to the host computer monitoring system through the bus;

(9) The upper computer will receive the temperature data signal, if the temperature exceeds the warning line, it will alarm, otherwise it will display the temperature;

(10), the host computer compares the real-time temperature with the forecast, if the deviation between the real-time data and the forecast data is too large, then combine the previous historical data samples to find a new forecast decision function, otherwise execute (11);

(11), predict and display the temperature data in the next period according to the decision function;

(12), store the data sample, and jump to (4);

The present invention performs on-line monitoring on the real-time temperature and on-line prediction on the temperature at the next moment. The system uses a thermoelectric temperature sensor for temperature collection, and the temperature signal is transmitted to the terminal monitoring equipment through wireless. This method can well solve the problem of high-voltage insulation between the high and low voltage sides of the temperature measurement system of the switchgear, without wiring, and the cost is relatively low. Low. At the same time, the support vector machine algorithm with stable performance is used for predictive modeling, which can quickly, accurately and robustly predict the temperature of the next period, and better play an early warning role. The invention has a very good market prospect.

The foregoing descriptions of specific exemplary embodiments of the present invention have been presented for purposes of illustration and description. These descriptions are not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling others skilled in the art to make and use various exemplary embodiments of the invention, as well as various Choose and change. It is intended that the scope of the invention be defined by the claims and their equivalents.

Claims (2)

1. wireless on-line monitoring system framework

(1), initiation parameter；

(2), host computer and receiving terminal microprocessor serial ports address configuration；

(3), obtain forecast and decision function according to conventional historical data sample, it was predicted that subsequent period temperature, and show；

(4), temperature acquisition transmitting terminal temperature sensor collecting temperature data be converted into digital signal；

(5), temperature acquisition transmitting terminal microprocessor carries out packing process to data；

(6), temperature acquisition transmitting terminal radio-frequency module delivers a packet to receiving terminal；

(7), receiving terminal Receiver Module receives the wireless signal from same geocoding, and stores the data in depositor；

(8), receiving terminal microprocessor extract the data in depositor, transfer data to Monitor Computer Control System by bus；

(9), host computer receives temperature data signal, if temperature exceedes warning line warning, otherwise displays temperature；

(10), real time temperature compared with prediction, if real time data is excessive with prediction data deviation, is again obtained new forecast and decision function, is otherwise performed (11) by host computer in conjunction with conventional historical data sample；

(11), predict the temperature data of subsequent period according to decision function and show；

(12), store data sample, and redirect (4).

2. the SVM prediction algorithm of wireless temperature online detection system.

(1), initiation parameter；

(2), read, update historical data sample；

(3), data normalization, smoothing, quantification treatment；

(4), training, forecast sample are formed；

(5), according to Sample Establishing object function；

(6) Algorithm for Solving object function, is utilized to obtain decision function

(7), utilize forecast sample that subsequent period is predicted

(8), judge the real-time sample difference with forecast sample, if excessive (exceeding threshold value) redirects (2), otherwise redirect (7).