CN102722166A - Intelligent vision detection system and state detection method of transformer substation device - Google Patents

Abstract

The invention discloses a substation inspection robot vision system and a state detection method, which includes a motion control module, the motion control module is provided with a pan-tilt control module, and the pan-tilt control module is connected with an image acquisition module through a visual servo module. The image acquisition module is connected with the image intelligent analysis system through the image transmission module. The image intelligent analysis system includes the equipment classification module, the knife switch status identification module, the high-voltage circuit breaker identification module, the instrument reading module and the abnormal appearance identification module. The module is connected with the comprehensive analysis module, and the comprehensive analysis module is connected with the substation information control system interface module. The present invention automatically analyzes the current state of the equipment, covering all the detection content of the daily inspection work of the substation equipment; realizes the transition from the periodic inspection mode to the state-based inspection mode; completes the automatic identification and recording of the status of the substation equipment, greatly reducing the number of staff participation.

Description

Intelligent visual detection system and state detection method for substation equipment

the

technical field

The present invention relates to an intelligent visual detection system and detection method. In particular, it relates to an intelligent visual detection system and a state detection method for substation equipment.

Background technique

In recent years, with the rapid and continuous development of China's national economy, electricity is the basis for the development of various industries, and the length of transmission lines and the number of substations have followed a sharp increase. The traditional power equipment inspection work relies on manual inspection of equipment status on a regular basis to ensure the daily safe operation of substations. With the increasing power consumption and the number of substations, the contradictions such as low efficiency of manual inspection, scattered detection quality, and no connection with the existing information system are becoming more and more obvious. Obviously, manual inspection can no longer meet the needs of the development of the times.

Especially after special geographical conditions such as plateaus and Gobi or extreme weather such as strong winds, fog, ice and snow, hail, thunderstorms, etc., manual inspections cannot be performed in time, and there are greater safety risks in inspections due to limited conditions. At this time, the substation In the event of an accident, decision makers will often delay the best time to deal with the accident because they cannot obtain intuitive on-site information in time, causing unnecessary economic losses and even casualties and other major accidents. The automation and modernization of power equipment operation and maintenance has increasingly shown its urgency.

With the development of science and technology, the construction technology of intelligent substation is becoming more and more mature, and unattended substation becomes possible. Electric robots can replace or assist manual inspections, improve work efficiency, reduce labor intensity, and reduce operational risks. It has become one of the normative contents of the State Grid for the construction of smart substations.

As an important part of the robot, the visual inspection system occupies a very important position in the process of replacing manual inspection. The visual inspection system is mainly based on visual sensors, such as CCD, CMOS, etc., to obtain image information of power equipment, and to detect the current state of the equipment by processing and analyzing the image.

The existing visual inspection system based on substation inspection robot has the following characteristics:

(1) Single function. Existing visual inspection systems for power equipment often only have the detection functions of a few special equipments, such as Chinese Patent Application No. Infrared detection method for substation knife switch pattern recognition, and other state detection methods specifically for circuit breakers and knife switches. Obviously, the state detection of the above-mentioned special equipment cannot satisfy the detection of the state of all equipment in the substation, and then cannot completely replace the function of manual inspection.

(2) Judging the status of the device only by the detection result of a single image has poor reliability. The existing visual inspection module only gives the detection result of the equipment status of a single image, and does not refer to the inspection method of confirming the equipment status from multiple angles during manual inspection. Therefore, the results of image detection directly affect the reliability of robot state detection, and due to various factors such as image quality and lighting effects in outdoor environments, it is difficult to achieve complete reliability of image detection. Therefore, relying solely on the results of single image processing, the reliability of the existing visual inspection system cannot be guaranteed.

(3) The image information collection method is highly dependent on the positioning accuracy of the robot. Most of the existing visual detection methods use fixed point detection, that is, the robot will stop at a fixed position every time it inspects, and set the pan/tilt to a fixed position to collect equipment images. And other factors, the target device cannot be guaranteed to appear accurately within the image field of view, which is an important factor leading to the failure of visual inspection.

(4) Lack of intelligent analysis function. Most of the existing visual inspection systems use status inspection. After the status inspection is completed, it is still necessary to manually judge whether the status is correct based on the work ticket information and historical records, and the degree of integration with the substation information management system is not high, and the on-site inspection results cannot be sent to the information management system. automatic update of .

Due to the above-mentioned problems of the existing visual detection system, it is difficult to realize reliable automatic detection of the state of all equipment in the substation, and the detection accuracy and stability need to be improved, so that the reliability of the detection of the state of the substation equipment by robots cannot be guaranteed, thus affecting the substation Hidden dangers have been buried in safe operation. How to solve the above defects and realize a set of visual detection system with automatic detection function of all substation equipment, and with high reliability and high precision has become a very urgent issue.

Contents of the invention

The purpose of the present invention is to overcome the incompleteness of the above-mentioned prior art, and provide an intelligent visual detection system and detection method for substation equipment based on a substation inspection robot platform. It has comprehensive functions, covering the automatic inspection function of all important substation equipment; it has a visual servo function, which does not depend on the system accuracy of the pan/tilt and mobile equipment; it can also conduct comprehensive analysis based on historical information and related equipment status, and directly give alarm information , without the need for staff to inspect substation equipment through real-time monitoring, manual recording, etc.

To achieve the above object, the present invention adopts the following technical solutions:

A substation inspection robot vision system, which includes a motion control module, the motion control module is provided with a pan-tilt control module, the pan-tilt control module is connected to an image acquisition module through a visual servo module, and the image acquisition module communicates with an image acquisition module through an image transmission module The image intelligent analysis system is connected. The image intelligent analysis system includes the equipment classification module, the knife switch state identification module, the high voltage circuit breaker identification module, the instrument reading module and the abnormal appearance identification module. The modules of the image intelligent analysis system are connected with the comprehensive analysis module. The analysis module is connected with the interface module of the substation information control system.

The image intelligent analysis system automatically identifies the equipment type through the equipment type identification module, and inputs the image into the corresponding identification module; after the state detection of the corresponding identification module is completed, the result is input into the comprehensive analysis module, and compared with the historical data Analyze and compare the status of related equipment to obtain the final detection results and early warning information. Finally, the equipment status information and the judgment result of whether the status is normal are provided to the substation staff through the substation information system interface module and the substation's own information system.

The instrument reading module adopts a quasi-circular pointer instrument reading method for the instrument image acquired by the image signal acquisition device, automatically extracts the quasi-circular area, extracts the position of the instrument pointer, and gives the current index of the instrument.

The knife switch state identification module processes and analyzes the confidence of the knife switch image acquired by the image signal acquisition device, and automatically recognizes the three situations of the knife switch being closed, disconnected, and not in place;

The high-voltage circuit breaker recognition module recognizes the circuit breaker image obtained by the image signal acquisition device, and mainly uses a method based on template matching to judge the opening and closing state of the circuit breaker;

The abnormal appearance identification module mainly detects hanging objects on the equipment, discharge traces and oil stains on the surface of the equipment.

The motion control module also includes a mobile platform and a motion controller that controls the movement of the mobile platform; the cloud platform control module includes a cloud platform and a cloud platform controller, and the cloud platform controller controls the rotation angle of the platform; the visual servo module utilizes The current image acquired by the image signal acquisition device is compared and analyzed with the pre-acquired template image to obtain the adjustment amount of the pan/tilt, and then adjust the attitude of the pan/tilt to ensure that the equipment image appears completely in the image and provide sufficient information for subsequent equipment status identification; The image transmission module includes a wireless router and a signal receiving base station, and transmits the compressed and coded image information to the background processing module through the wireless network.

A state detection method using the vision system of a substation inspection robot, the steps are as follows:

(1) Control the motion control platform based on the substation inspection robot to move to the designated position, adjust the image acquisition module through the visual servo module to ensure that the equipment image appears in the image completely, and then obtain a clear image of the equipment to be detected through the image acquisition module;

(2) Compress and encode the collected image information first, and then transmit it to the image intelligent analysis system through the image transmission module;

(3) The equipment type identification module of the image intelligent analysis system classifies the images according to the knife switch, high-voltage circuit breaker, and instrument, and automatically sends the equipment image to the corresponding identification module;

(4) The knife switch state identification module, high-voltage circuit breaker identification module, instrument reading module, and transformer appearance abnormality identification module respectively conduct consistency analysis on their current images and associated equipment historical images through the comprehensive analysis module, and finally the equipment status The information and early warning information are transmitted to the substation information system interface module;

(5) The substation information system interface module converts the detection and analysis results into the data type recognized by the substation information system, and transmits them to the substation information management system of the upper computer;

(6) The staff directly monitors the judgment results.

In the step (3), the automatic classification of the device classification module mainly obtains the current device type to be detected through the configuration information of the machine docking position, so as to automatically determine the detection module to be entered. The method is simple, reliable and easy to implement.

In the step (4), the knife switch state identification module judges the knife switch state, and automatically recognizes the three situations of the knife switch being closed, disconnected, and not in place;

The high-voltage circuit breaker identification module judges the state of the high-voltage circuit breaker, and judges the opening and closing state of the high-voltage circuit breaker;

The meter reading module judges the state of the meter, automatically recognizes the pointer position of the meter, and gives the current index of the meter;

The abnormal appearance identification module identifies whether the equipment has hanging objects, discharge traces, oil stains, etc. on the surface of the equipment.

The specific process of the meter reading module judging the state of the meter is:

(1) Grayscale and binarize the instrument image collected by the mobile robot;

(2) Use the Canny algorithm to extract the edge information of the binary image of the device; use the edge points to fit the ellipse using the least squares method; filter the obtained ellipse, remove the non-ellipse image that is fitted into an ellipse, and obtain the dial area and dial Center position;

(3) Use the straight line extraction algorithm based on Hough transform to extract the pointer in the obtained dial image area, refer to the center of the dial, and then obtain the angle of the pointer;

(4) Extract the relationship between the pointer angle and the indication corresponding to the target instrument in the database and calculate the indication corresponding to the angle obtained in step (3).

In the step (4), the analysis process of the comprehensive analysis module is: using the substation equipment distribution information, automatically record and save the status of the associated equipment, and perform consistency analysis; through this module, the current equipment status is longitudinally Compared with the historical state, the state is compared with the connected equipment horizontally, the state of the equipment is comprehensively analyzed, and the analysis results are given, and the equipment with abnormal state is given an alarm prompt. The specific steps are as follows:

(1) According to the daily inspection tasks of the substation, associate the inspection content of the associated equipment;

(2) According to the state sequence control relationship of related equipment, after the execution of the associated terminal task is completed, the consistency of each equipment in the associated task is judged;

(3) For the status of abnormal equipment, compare it with the historical data of the previous two inspections, and combine the sequence of work tickets to conduct consistency analysis to determine whether an abnormality occurs.

Beneficial effects of the present invention:

(1) For the first time, the system covers the status detection function of all inspection equipment in substations, and proposes an automatic detection algorithm for instruments, an automatic identification algorithm for equipment types, automatic identification of equipment types, and truly realizes fully automated detection of various types of equipment in substations. device status;

(2) An image-based visual servo algorithm is proposed, which automatically realizes the automatic servo of the pan/tilt according to the position of the target image, ensures that the target image appears completely within the field of view, and improves the overall detection accuracy;

(3) A device state consistency inspection algorithm that automatically correlates detection results with historical data and associated equipment state is proposed, which increases the reliability of state detection, and directly gives alarm information, and truly realizes the transformation to the state-based inspection method ;

(4) Based on the equipment status comprehensive analysis module proposed by this intelligent visual inspection system, the daily inspection tasks of substation equipment can be completely replaced by robots. Improve work efficiency;

(5) Based on the intelligent visual inspection system, the substation inspection robot has a high-precision, high-reliability automatic inspection method, which provides a solid technical guarantee for the safe operation of the substation and reduces safety accidents, and provides a safe and stable development for the power industry Foundation.

Description of drawings

Figure 1 is a schematic structural diagram of an intelligent visual inspection system for substation equipment.

Among them, 1. Image acquisition module, 2. PTZ control module, 3. Visual servo module, 4. Image transmission module, 5. Image intelligent analysis system, 6. Equipment type identification module, 7. Knife gate status identification module, 8 .High-voltage circuit breaker identification module, 9. Instrument reading module, 10. Abnormal appearance identification module, 11. Comprehensive analysis module, 12. Substation information system interface module, 13. Motion control module.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

As shown in Figure 1, an intelligent visual detection system for a substation inspection robot includes an image acquisition module 1, a pan/tilt control module 2, a visual servo module 3, an image transmission module 4, a comprehensive analysis module 11, and a substation information system interface module 12 , motion control module 13; Wherein, motion control module 13 is provided with pan-tilt control module 2, and pan-tilt control module 2 is connected with image acquisition module 1 through visual servo module 3, and image acquisition module 1 is connected with image transmission module 4, image The transmission module 4 is connected with the intelligent image analysis system 5 , the intelligent image analysis system 5 is connected with the comprehensive analysis module 11 , and the comprehensive analysis module 11 is connected with the substation information system interface 12 . After the image information is collected by the image acquisition module 1 , it is connected with the image intelligent analysis system 5 through the image transmission module 4 . The image intelligent analysis system 5 includes an equipment type identification module 6 , a knife switch state identification module 7 , a high voltage circuit breaker identification module 8 , an instrument reading module 9 and an abnormal appearance identification module 10 .

The image acquisition module 1 includes an image signal acquisition device and an image codec device. The image signal acquisition device can be a monocular camera or a CCD industrial camera, and the image encoding device is responsible for compressing and encoding the image analog signal.

The motion control module 13 includes a mobile platform and a motion controller for controlling the movement of the mobile platform.

The pan-tilt control module 2 includes a pan-tilt and a pan-tilt controller, and the pan-tilt controller can set, save, and control the rotation angle of the pan-tilt. The image signal acquisition device in the image acquisition module 1 is mounted on the pan-tilt, and the pan-tilt controller controls the horizontal and vertical rotation angles of the pan-tilt, thereby controlling the attitude of the camera and flexibly shooting different devices from multiple angles.

The visual servo module 3 compares and analyzes the current image acquired by the image signal acquisition device with the pre-collected template image to obtain the adjustment amount of the pan/tilt, and then adjusts the attitude of the pan/tilt to ensure that the image of the equipment is completely present in the image, and is used for subsequent equipment. Status recognition provides sufficient information. The structure adopted by the system adopts the structure described in the Chinese authorized patent number ZL201020685635.7. Based on the visual servo function, when the substation inspection robot reaches a fixed detection point, it can independently adjust the pan/tilt according to the current image information, so that the target image appears completely within the field of view. The situation that the target deviates from the field of view caused by the system error of the pan-tilt and the mobile platform is eliminated, and the detection accuracy and reliability are improved.

The image transmission module 4 includes a wireless router and a signal receiving base station, and transmits the compressed and coded image information to the background processing module through a wireless network. The structure adopted by the system adopts the wireless communication module structure described in the Chinese authorized patent number 201110005671.3.

The image intelligent analysis system 5 automatically identifies equipment types through the equipment type identification module 6, such as images of knife switches, high-voltage circuit breakers, instruments, etc., and automatically inputs the images into the corresponding identification modules. After the status detection of the corresponding identification module is completed, the result is input into the comprehensive analysis module 11, and the final detection result and early warning information are obtained through comparative analysis with historical data and related equipment status. Finally, the equipment state information and the judgment result of whether the state is normal are provided to the substation staff through the substation information system interface module 12 and the substation's own information system.

The device type recognition module 6 automatically recognizes the device type, and obtains the current device type to be detected mainly through the configuration information of the machine parking position, thereby automatically determining the detection module to be entered.

The knife switch state identification module 7 processes and analyzes the knife switch image acquired by the image signal acquisition device, and automatically recognizes the three situations of the knife switch being closed, disconnected, and not closed properly. The structure adopted by the system adopts the structure and method described in Chinese Authorized Patent No. 2011102163781.

The high-voltage circuit breaker recognition module 8 recognizes the circuit breaker image obtained by the image signal acquisition device, and mainly uses a method based on template matching to judge the opening and closing state of the circuit breaker. The structure adopted by the system adopts the structure and method described in Chinese Authorized Patent No. 201110216396X.

The instrument reading module 9 adopts a quasi-circular pointer instrument reading method for the instrument image acquired by the image signal acquisition device, automatically extracts the quasi-circular area, extracts the position of the instrument pointer, and gives the current index of the instrument. This module manually extracts the template of the target instrument, marks the corresponding relationship between the scale and the angle of the target instrument, and stores it in the database. The specific method includes the following steps:

(1) Grayscale and binarize the instrument image collected by the mobile robot;

(2) Use the Canny algorithm to extract the edge information of the binary image of the device; use the edge points to fit the ellipse using the least squares method; filter the obtained ellipse, remove the non-ellipse image that is fitted into an ellipse, and obtain the dial area and dial Center position;

(3) Use the straight line extraction algorithm based on Hough transform to extract the pointer in the obtained dial image area, refer to the center of the dial, and then obtain the angle of the pointer;

(4) Extract the relationship between the pointer angle and the indication corresponding to the target instrument in the database and calculate the indication corresponding to the angle obtained in step (3).

This method is based on the existing inspection robot detection method, based on fixed detection points, and adopts the visual servo function at the same time, which greatly improves the reliability and flexibility of instrument identification, and is superior to existing instrument detection methods.

The abnormal appearance identification module 10 refers to detecting abnormalities on the surface of equipment such as hanging objects, discharge traces, oil stains, etc. by analyzing the visible information on the surface of the equipment. One of the important content. The structure adopted by the system adopts the structure and method described in the Chinese authorized patent number ZL 201020555312.6.

The comprehensive analysis module 11 uses the substation equipment distribution information to automatically record and save the state of the associated equipment, and perform consistency analysis. Through this module, the current equipment status can be compared vertically with the historical status, and horizontally compared with the status of the connected equipment, comprehensively analyze the status of the equipment, and give the analysis results, and give an alarm prompt for the equipment with abnormal status. The specific implementation method of this module:

(1) According to the daily inspection tasks of the substation, associate the inspection content of the associated equipment,

(2) According to the state sequence control relationship of related equipment, after the execution of the associated terminal task is completed, the consistency of each equipment in the associated task is judged;

(3) For the status of abnormal equipment, compare it with the historical data of the previous two inspections, and combine the sequence of work tickets to conduct consistency analysis to determine whether an abnormality occurs.

The substation information system interface module 12 connects the detected equipment status and early warning information to the substation MIS system, and the substation staff can directly understand the current status of various equipment in the substation through the information system in the substation, which can completely replace manual inspections. Examining equipment abnormalities only based on early warning information will greatly reduce the labor intensity of staff.

The specific implementation process of the substation equipment status detection method using the above system is as follows:

(1) Based on the motion control platform of the substation inspection robot moving to a specific position, the clear image of the equipment to be inspected is obtained through the image acquisition module 1, and the image acquisition module 1 is adjusted through the visual servo module 3 to ensure that the complete image of the equipment appears in the image middle;

(2) Compress and encode the image information obtained, and then transmit it to the background processor through the image transmission module 4;

(3) Use the device type identification module to automatically classify the device type, and input the image to the corresponding device status detection module;

(4) The corresponding equipment status detection system transmits the detection results to the comprehensive analysis module 11, which automatically analyzes the consistency of the current equipment status and historical data with the associated equipment, and finally transmits the equipment status information and early warning information to the substation Information system interface module 12;

(5) The substation information system interface module 12 converts the detection and analysis results into the types identified by the substation information system, and transmits them to the substation information management system;

(6) The staff directly monitors the judgment results.

Similar processes are also applicable to the identification of high-voltage circuit breakers, instrument readings, and identification of abnormal appearance of equipment.

Although the specific implementation of the present invention has been described above in conjunction with the accompanying drawings, it does not limit the protection scope of the present invention. Those skilled in the art should understand that on the basis of the technical solution of the present invention, those skilled in the art do not need to pay creative work Various modifications or variations that can be made are still within the protection scope of the present invention.

Claims (10)

1. substation inspection robotic vision system; It is characterized in that; It comprises motion-control module, and said motion-control module is provided with the cradle head control module, and the cradle head control module links to each other with image capture module through the visual servo module; Image capture module links to each other with the image intelligent analytic system through image transmission module; The image intelligent analytic system comprises device class module, disconnecting link state recognition module, primary cut-out identification module, meter reading module and the unusual identification module of outward appearance, and each module of image intelligent analytic system is connected with the analysis-by-synthesis module, and the analysis-by-synthesis module links to each other with transformer station management information system interface module.

2. substation inspection robotic vision system as claimed in claim 1 is characterized in that, said image intelligent analytic system is passed through device type identification module automatic identification equipment type, and image is input in the corresponding identification module; After corresponding identification module state-detection is accomplished; The result is input in the analysis-by-synthesis module; Through with historical data comparative analysis, the comparative analysis of associate device state, obtain final detection result and early warning information, at last with status information of equipment and state normal judged result whether; In the own infosystem of transformer station's infosystem interface module and transformer station, supply the staff of transformer station to use.

3. substation inspection robotic vision system as claimed in claim 1; It is characterized in that; The meter diagram picture that said meter reading module is obtained the picture signal collecting device adopts similar round readings of pointer type meters method; Automatically extract the similar round zone, extraction apparatus list index position provides the instrument current time index.

4. substation inspection robotic vision system according to claim 1; It is characterized in that; Said disconnecting link state recognition module is handled and is analyzed the disconnecting link picture signal that the picture signal collecting device obtains, and discerns closure, disconnection, closed three kinds of situation not in place of disconnecting link automatically;

Said primary cut-out identification module is discerned through the isolating switch image that the picture signal collecting device obtains, utilize based on the method for template matches judge isolating switch branch, close state;

The unusual identification module of said outward appearance, to have hanger on the equipment, equipment surface exists spark tracking, oil stain class situation to detect.

5. substation inspection robotic vision system as claimed in claim 1 is characterized in that, said motion-control module also comprises the motion controller that mobile platform and control mobile platform move; Said cradle head control module comprises The Cloud Terrace and cradle head control device, cradle head control device control The Cloud Terrace rotational angle; Said visual servo module utilizes picture signal collecting device present image that obtains and the template image of gathering in advance to compare analysis; Obtain the The Cloud Terrace adjustment amount; Then adjust the The Cloud Terrace attitude; Guarantee that equipment drawing appears in the image as complete, for follow-up equipment state identification provides sufficient information; Image transmission module comprises that wireless router, signal receive the base station, pass through wireless network transmissions in background processing module with the image information behind the compressed encoding.

6. condition detection method that utilizes the substation inspection robotic vision system of claim 1 is characterized in that step is following:

(1) control is controlled platform based on the substation inspection robot motion and is moved to assigned address; Through the visual servo module image capture module is adjusted to guarantee that equipment drawing appears in the image as complete, obtained the picture rich in detail of equipment to be detected then through image capture module;

(2) with the image information that collects, carry out compressed encoding earlier, be transferred in the image intelligent analytic system through image transmission module again;

(3) the device type identification module of image intelligent analytic system is discerned image according to disconnecting link, primary cut-out, instrument and transformer, and automatically the image of equipment is sent into corresponding identification module;

(4) disconnecting link state recognition module, primary cut-out identification module, meter reading module and the unusual identification module of outward appearance carry out consistency analysis with separately present image and associated device history image through the analysis-by-synthesis module respectively, and status information of equipment and early warning information transfer to transformer station's infosystem interface module the most at last;

(5) transformer station's infosystem interface module will detect with analysis result and convert transformer station's infosystem recognition data type to, transfer in transformer station's information management system of host computer;

(6) staff directly monitors judged result.

7. condition detection method as claimed in claim 6; It is characterized in that in the said step (3), the configuration information of machine stop position is mainly passed through in the automatic classification of device class module; Obtain current device type to be detected, thus definite automatically detection module that will get into.

8. condition detection method as claimed in claim 6 is characterized in that, in the said step (4), disconnecting link state recognition module is judged the disconnecting link state, discerns closure, disconnection, closed three kinds of situation not in place of disconnecting link automatically;

The primary cut-out identification module is judged high-voltage circuit-breaker status, judge primary cut-out branch, close state;

The meter reading module is judged meter status, discerns the pointer position of instrument automatically, provides the instrument current time index;

Whether the unusual identification module identification equipment of outward appearance has hanger, equipment surface exists spark tracking, oil stain class situation outward appearance ERST.

9. condition detection method as claimed in claim 8 is characterized in that, said meter reading module judges that the detailed process of meter status is:

(1) the instrument image gray processing of gathering the mobile robot, binaryzation;

(2) utilize the Canny algorithm to extract the marginal information of equipment bianry image; Utilize marginal point to use least square fitting oval; Ellipse to obtaining filters, and weeds out to be fitted to oval non-oval image, obtains dial plate zone and dial plate home position;

(3) in obtaining the dial plate image-region, use straight line extraction algorithm to extract pointer,, and then obtain the angle of pointer with reference to the dial plate center of circle based on the Hough conversion;

(4) concern the corresponding registration of calculation procedure (3) gained angle between the pointer angle of extracting corresponding target instrument in the database and the registration.

10. condition detection method as claimed in claim 6 is characterized in that, in the said step (4); The analytic process of analysis-by-synthesis module is: utilize the substation equipment distributed intelligence; Automatically with the state recording of associated device, preserve, and carry out consistency analysis; Through this module, the current device state is contrasted with historic state longitudinally, horizontal carries out the state comparison with associated devices; The state of this equipment of analysis-by-synthesis; And provide analysis result, and carry out alarm for the equipment that abnormal state occurs, its concrete steps are:

(1) according to the daily patrol task of transformer station, associated device is maked an inspection tour content carry out association,

(2), after the terminal task of association is complete, the consistance of each equipment in the associated task is judged according to relevant device sequence of states control relation;

(3) for producing unusual equipment state, compare with the historical data of preceding secondary detection, combine the work ticket order, carry out consistency analysis, determine whether to take place unusually.